--------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  /Users/tobiasrommel/Downloads/Rommel_FDIsupport_CPS_Replication/Rommel_FDIsupport_CPS_Analysis.log
  log type:  text
 opened on:  28 Jun 2023, 11:36:39

. 
. 
. 
. ***** 1. Afrobarometer Survey Data
. *------------------------------------------------------------------------------*
. 
. *** Load data
. use "Rommel_FDIsupport_CPS_AFB.dta", clear

. 
. *** TABLE A-2: Share of respondents exposed to FDI
. tab cname, sum(fdi_treat)

            |  Summary of Potentially exposed to
            |                 FDI
    Country |        Mean   Std. dev.       Freq.
------------+------------------------------------
    Algeria |    .4030782   .49061828       2,404
  Burkina F |   .15111111    .3582069       3,600
   Cameroon |   .29806885   .45750594       2,382
      Egypt |   .47529313   .49949379       2,388
      Gabon |   .57929883   .49387782       1,198
     Guinea |         .13   .33637352       2,400
  Ivory Coa |   .33305544   .47140444       2,399
  Madagasca |   .25653333   .43677781       3,750
       Mali |   .16299559   .36941251       3,632
    Morocco |   .53881469    .4985952       2,396
  Mozambiqu |   .29066667   .45410783       6,000
    Namibia |   .38888889   .48756574       3,600
    Nigeria |   .29309377   .45521303       7,124
      Sudan |   .09337224   .29101403       2,399
  Swaziland |         .36   .48010003       2,400
   Tanzania |   .25642309   .43669452       5,994
       Togo |         .35     .477069       2,400
    Tunisia |   .62833333    .4833507       2,400
     Uganda |   .32955331   .47008349       7,231
   Zimbabwe |   .29883333   .45778477       6,000
------------+------------------------------------
      Total |   .31053997   .46271791      72,097

. 
. *** Model specification
. keep if fdi_treat==1
(49,708 observations deleted)

. * Controls
. global C1 age female unempl urban

. global C2 interest news 

. global C3 i.race 

. * Sample
. xtreg problem c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,644
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0110                                         min =        218
     Between = 0.0209                                         avg =    1,082.2
     Overall = 0.0074                                         max =      2,334

                                                Wald chi2(11)     =     238.82
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0134041   .0024538     5.46   0.000     .0085948    .0182135
               educlvl |   .0238654   .0034469     6.92   0.000     .0171096    .0306212
                       |
c.lnfdi_jobs#c.educlvl |  -.0030808   .0005086    -6.06   0.000    -.0040776   -.0020841
                       |
                   age |  -.0015513   .0002578    -6.02   0.000    -.0020565   -.0010461
                female |   .0029886   .0068484     0.44   0.663    -.0104339    .0164111
                unempl |  -.0288069   .0074873    -3.85   0.000    -.0434816   -.0141321
                 urban |   .0368757   .0084071     4.39   0.000     .0203981    .0533532
              interest |  -.0075861   .0033075    -2.29   0.022    -.0140687   -.0011035
                  news |   .0051705   .0025735     2.01   0.045     .0001265    .0102145
                       |
                 round |
                    5  |  -.0509178   .0104697    -4.86   0.000     -.071438   -.0303976
                    6  |  -.0613164   .0106789    -5.74   0.000    -.0822467   -.0403862
                       |
                 _cons |   .4914222   .0335938    14.63   0.000     .4255795    .5572649
-----------------------+----------------------------------------------------------------
               sigma_u |  .11847009
               sigma_e |  .48420753
                   rho |  .05648132   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. gen sample=e(sample)==1

. 
. *** TABLE A-1: Summary statistics, Afrobarometer
. tabstat problem condlive condecon trustelec trustlead trustpart demoqua demosat demosup ///
>         educlvl lnfdi_jobs lnfdi_capex age female unempl urban interest news if sample==1, ///
>         stats(N mean sd min max) format(%10.2f)

   Stats |    problem   condlive   condecon   truste~c   trustl~d   trustp~t    demoqua    demosat    demosup
---------+---------------------------------------------------------------------------------------------------
       N |   21644.00   21554.00   21375.00   18331.00   20584.00   19286.00   19887.00   20225.00   17407.00
    Mean |       0.51       3.22       3.46       0.56       0.45       0.59       2.60       1.80       0.84
      SD |       0.50       1.15       1.23       0.50       0.50       0.49       0.92       1.01       0.37
     Min |       0.00       1.00       1.00       0.00       0.00       0.00       1.00       0.00       0.00
     Max |       1.00       5.00       5.00       1.00       1.00       1.00       4.00       4.00       1.00
-------------------------------------------------------------------------------------------------------------

   Stats |    educlvl   lnfdi_~s   lnfdi_~x        age     female     unempl      urban   interest       news
---------+---------------------------------------------------------------------------------------------------
       N |   21644.00   21644.00   21644.00   21644.00   21644.00   21644.00   21644.00   21644.00   21644.00
    Mean |       4.15       5.43       4.60      35.58       0.50       0.32       0.71       1.64       1.54
      SD |       2.16       2.99       2.73      13.79       0.50       0.46       0.45       1.05       1.56
     Min |       0.00       0.00       0.00      18.00       0.00       0.00       0.00       0.00       0.00
     Max |       9.00      11.23      10.51     100.00       1.00       1.00       1.00       3.00       4.00
-------------------------------------------------------------------------------------------------------------

. 
. *** Set 1: Economic Grievances
. * Self-Reported Economic Problems
. xtreg problem c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,876
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0105                                         min =        221
     Between = 0.0208                                         avg =    1,093.8
     Overall = 0.0074                                         max =      2,356

                                                Wald chi2(9)      =     231.42
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0133728   .0024228     5.52   0.000     .0086242    .0181213
               educlvl |   .0243753   .0033381     7.30   0.000     .0178328    .0309178
                       |
c.lnfdi_jobs#c.educlvl |  -.0030547   .0005035    -6.07   0.000    -.0040416   -.0020678
                       |
                   age |  -.0016016   .0002552    -6.28   0.000    -.0021018   -.0011014
                female |   .0026962   .0067062     0.40   0.688    -.0104476    .0158401
                unempl |   -.030424   .0074361    -4.09   0.000    -.0449984   -.0158496
                 urban |   .0390676   .0082348     4.74   0.000     .0229277    .0552076
                       |
                 round |
                    5  |  -.0482873   .0104132    -4.64   0.000    -.0686968   -.0278777
                    6  |  -.0603487   .0106252    -5.68   0.000    -.0811738   -.0395237
                       |
                 _cons |    .483938   .0328704    14.72   0.000     .4195132    .5483627
-----------------------+----------------------------------------------------------------
               sigma_u |   .1166852
               sigma_e |  .48438407
                   rho |  .05484713   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store problem1

. xtreg problem c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,644
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0110                                         min =        218
     Between = 0.0209                                         avg =    1,082.2
     Overall = 0.0074                                         max =      2,334

                                                Wald chi2(11)     =     238.82
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0134041   .0024538     5.46   0.000     .0085948    .0182135
               educlvl |   .0238654   .0034469     6.92   0.000     .0171096    .0306212
                       |
c.lnfdi_jobs#c.educlvl |  -.0030808   .0005086    -6.06   0.000    -.0040776   -.0020841
                       |
                   age |  -.0015513   .0002578    -6.02   0.000    -.0020565   -.0010461
                female |   .0029886   .0068484     0.44   0.663    -.0104339    .0164111
                unempl |  -.0288069   .0074873    -3.85   0.000    -.0434816   -.0141321
                 urban |   .0368757   .0084071     4.39   0.000     .0203981    .0533532
              interest |  -.0075861   .0033075    -2.29   0.022    -.0140687   -.0011035
                  news |   .0051705   .0025735     2.01   0.045     .0001265    .0102145
                       |
                 round |
                    5  |  -.0509178   .0104697    -4.86   0.000     -.071438   -.0303976
                    6  |  -.0613164   .0106789    -5.74   0.000    -.0822467   -.0403862
                       |
                 _cons |   .4914222   .0335938    14.63   0.000     .4255795    .5572649
-----------------------+----------------------------------------------------------------
               sigma_u |  .11847009
               sigma_e |  .48420753
                   rho |  .05648132   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store problem2

. xtreg problem c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,550
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0110                                         min =        130
     Between = 0.0938                                         avg =    1,077.5
     Overall = 0.0130                                         max =      2,334

                                                Wald chi2(15)     =     240.81
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0133484   .0024568     5.43   0.000     .0085331    .0181638
               educlvl |   .0238599   .0034489     6.92   0.000     .0171001    .0306197
                       |
c.lnfdi_jobs#c.educlvl |  -.0030297   .0005095    -5.95   0.000    -.0040283   -.0020311
                       |
                   age |   -.001575   .0002585    -6.09   0.000    -.0020816   -.0010684
                female |   .0030602   .0068646     0.45   0.656    -.0103943    .0165146
                unempl |  -.0287061    .007506    -3.82   0.000    -.0434177   -.0139945
                 urban |   .0364319   .0084218     4.33   0.000     .0199254    .0529383
              interest |  -.0079642   .0033142    -2.40   0.016      -.01446   -.0014684
                  news |   .0048801   .0025809     1.89   0.059    -.0001784    .0099385
                       |
                  race |
                    2  |   .0405539   .0501545     0.81   0.419    -.0577471     .138855
                    3  |  -.0176689   .0328191    -0.54   0.590    -.0819932    .0466554
                    4  |   .0235719   .0365177     0.65   0.519    -.0480015    .0951452
                    5  |   -.002633   .0886719    -0.03   0.976    -.1764268    .1711608
                       |
                 round |
                    5  |  -.0509824   .0104703    -4.87   0.000    -.0715037   -.0304611
                    6  |  -.0616718   .0106876    -5.77   0.000    -.0826191   -.0407245
                       |
                 _cons |   .4897798   .0286746    17.08   0.000     .4335786     .545981
-----------------------+----------------------------------------------------------------
               sigma_u |  .08301583
               sigma_e |  .48393895
                   rho |  .02858546   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store problem3

. xtreg problem c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,644
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0111                                         min =        218
     Between = 0.0217                                         avg =    1,082.2
     Overall = 0.0076                                         max =      2,334

                                                Wald chi2(11)     =     240.95
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0153204   .0026969     5.68   0.000     .0100345    .0206062
                educlvl |    .023119    .003287     7.03   0.000     .0166766    .0295614
                        |
c.lnfdi_capex#c.educlvl |  -.0034542   .0005559    -6.21   0.000    -.0045438   -.0023646
                        |
                    age |  -.0015409   .0002578    -5.98   0.000    -.0020461   -.0010357
                 female |   .0029444    .006848     0.43   0.667    -.0104773    .0163662
                 unempl |  -.0287948    .007487    -3.85   0.000     -.043469   -.0141206
                  urban |   .0357077   .0084441     4.23   0.000     .0191576    .0522579
               interest |  -.0075194   .0033076    -2.27   0.023    -.0140022   -.0010366
                   news |   .0051126   .0025736     1.99   0.047     .0000684    .0101569
                        |
                  round |
                     5  |  -.0510476   .0104374    -4.89   0.000    -.0715046   -.0305907
                     6  |  -.0613736   .0106798    -5.75   0.000    -.0823056   -.0404417
                        |
                  _cons |   .4942375   .0332237    14.88   0.000     .4291202    .5593548
------------------------+----------------------------------------------------------------
                sigma_u |  .11789619
                sigma_e |  .48418375
                    rho |  .05597117   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store problem4

. xtprobit problem c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -14995.438  
Iteration 1:   log likelihood = -14869.996  
Iteration 2:   log likelihood = -14869.984  
Iteration 3:   log likelihood = -14869.984  

Fitting full model:

rho =  0.0     log likelihood = -14869.984
rho =  0.1     log likelihood = -14397.708
rho =  0.2     log likelihood =   -14452.2

Iteration 0:   log likelihood = -14385.026  
Iteration 1:   log likelihood = -14346.585  
Iteration 2:   log likelihood = -14346.579  
Iteration 3:   log likelihood = -14346.579  

Random-effects probit regression                    Number of obs    =  21,644
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     218
                                                                 avg = 1,082.2
                                                                 max =   2,334

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  232.96
Log likelihood = -14346.579                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0356572   .0065493     5.44   0.000     .0228208    .0484935
               educlvl |    .063722   .0091895     6.93   0.000     .0457108    .0817332
                       |
c.lnfdi_jobs#c.educlvl |  -.0082417   .0013572    -6.07   0.000    -.0109017   -.0055816
                       |
                   age |  -.0040879   .0006813    -6.00   0.000    -.0054231   -.0027526
                female |   .0077719   .0180642     0.43   0.667    -.0276333     .043177
                unempl |   -.075142    .019769    -3.80   0.000    -.1138885   -.0363956
                 urban |   .0958589   .0222114     4.32   0.000     .0523253    .1393925
              interest |  -.0196636   .0087148    -2.26   0.024    -.0367442    -.002583
                  news |   .0131094   .0067791     1.93   0.053    -.0001773    .0263962
                       |
                 round |
                    5  |  -.1311325    .027481    -4.77   0.000    -.1849942   -.0772708
                    6  |  -.1585194   .0280467    -5.65   0.000    -.2134899   -.1035489
                       |
                 _cons |  -.0277421   .0914861    -0.30   0.762    -.2070515    .1515673
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.231733   .3237187                      -2.86621   -1.597256
-----------------------+----------------------------------------------------------------
               sigma_u |   .3276313   .0530302                       .238567    .4499459
                   rho |   .0969368   .0283384                      .0538494    .1683655
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 1046.81                Prob >= chibar2 = 0.000

. est store problem5

. xtreg problem c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,644
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0102                                         min =        218
     Between = 0.0198                                         avg =    1,082.2
     Overall = 0.0068                                         max =      2,334

                                                Wald chi2(11)     =     221.62
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                  problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0089776   .0021158     4.24   0.000     .0048308    .0131245
                  educlvl |   .0141009     .00252     5.60   0.000     .0091618      .01904
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0019526   .0004285    -4.56   0.000    -.0027924   -.0011128
                          |
                      age |  -.0015053   .0002579    -5.84   0.000    -.0020109   -.0009998
                   female |   .0025642   .0068525     0.37   0.708    -.0108665     .015995
                   unempl |  -.0286156    .007492    -3.82   0.000    -.0432995   -.0139316
                    urban |   .0365957    .008634     4.24   0.000     .0196735     .053518
                 interest |  -.0077569   .0033078    -2.35   0.019    -.0142401   -.0012737
                     news |   .0053822   .0025708     2.09   0.036     .0003435    .0104209
                          |
                    round |
                       5  |  -.0485637   .0103257    -4.70   0.000    -.0688017   -.0283258
                       6  |  -.0592457    .010408    -5.69   0.000    -.0796451   -.0388463
                          |
                    _cons |   .5285359   .0271901    19.44   0.000     .4752444    .5818274
--------------------------+----------------------------------------------------------------
                  sigma_u |  .08903734
                  sigma_e |  .48438719
                      rho |  .03268344   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store problem6

. * Currently Bad Living Conditions
. xtreg condlive c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     22,094
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0373                                         min =        222
     Between = 0.1644                                         avg =    1,104.7
     Overall = 0.0448                                         max =      2,361

                                                Wald chi2(9)      =     857.82
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0102093   .0054565     1.87   0.061    -.0004852    .0209038
               educlvl |  -.0653799   .0075322    -8.68   0.000    -.0801426   -.0506171
                       |
c.lnfdi_jobs#c.educlvl |  -.0012598   .0011353    -1.11   0.267     -.003485    .0009654
                       |
                   age |   .0048437   .0005746     8.43   0.000     .0037175    .0059698
                female |  -.0394686   .0151211    -2.61   0.009    -.0691054   -.0098318
                unempl |  -.1409123   .0167563    -8.41   0.000    -.1737541   -.1080706
                 urban |  -.0895308   .0185777    -4.82   0.000    -.1259424   -.0531192
                       |
                 round |
                    5  |   .1525276   .0234849     6.49   0.000     .1064981    .1985571
                    6  |  -.0782974   .0239479    -3.27   0.001    -.1252345   -.0313604
                       |
                 _cons |   3.394583   .0749967    45.26   0.000     3.247592    3.541574
-----------------------+----------------------------------------------------------------
               sigma_u |  .26827902
               sigma_e |  1.0976081
                   rho |  .05637399   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store condlive1

. xtreg condlive c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0442                                         min =        219
     Between = 0.2551                                         avg =    1,093.0
     Overall = 0.0532                                         max =      2,337

                                                Wald chi2(11)     =    1013.24
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0155547   .0055124     2.82   0.005     .0047505    .0263589
               educlvl |  -.0434118   .0077559    -5.60   0.000     -.058613   -.0282105
                       |
c.lnfdi_jobs#c.educlvl |  -.0019371   .0011436    -1.69   0.090    -.0041785    .0003043
                       |
                   age |   .0049714   .0005786     8.59   0.000     .0038373    .0061055
                female |  -.0685286   .0153908    -4.45   0.000    -.0986939   -.0383633
                unempl |  -.1504654   .0168174    -8.95   0.000     -.183427   -.1175038
                 urban |   -.048071   .0188974    -2.54   0.011    -.0851092   -.0110329
              interest |  -.0261323   .0074271    -3.52   0.000    -.0406891   -.0115755
                  news |  -.0662195   .0057804   -11.46   0.000     -.077549   -.0548901
                       |
                 round |
                    5  |   .1470157   .0235374     6.25   0.000     .1008832    .1931481
                    6  |  -.0843234   .0239895    -3.52   0.000     -.131342   -.0373047
                       |
                 _cons |    3.41868   .0716359    47.72   0.000     3.278277    3.559084
-----------------------+----------------------------------------------------------------
               sigma_u |  .24414612
               sigma_e |  1.0937035
                   rho |  .04746583   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store condlive2

. xtreg condlive c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,763
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0455                                         min =        130
     Between = 0.2233                                         avg =    1,088.2
     Overall = 0.0532                                         max =      2,337

                                                Wald chi2(15)     =    1039.90
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0153668   .0055137     2.79   0.005     .0045601    .0261735
               educlvl |  -.0434351   .0077514    -5.60   0.000    -.0586277   -.0282426
                       |
c.lnfdi_jobs#c.educlvl |  -.0017379   .0011443    -1.52   0.129    -.0039806    .0005048
                       |
                   age |   .0051744   .0005795     8.93   0.000     .0040386    .0063103
                female |  -.0666027   .0154053    -4.32   0.000    -.0967966   -.0364087
                unempl |  -.1489336    .016838    -8.85   0.000    -.1819355   -.1159317
                 urban |  -.0469839   .0189293    -2.48   0.013    -.0840847   -.0098831
              interest |  -.0268507   .0074341    -3.61   0.000    -.0414213   -.0122802
                  news |  -.0648252   .0057956   -11.19   0.000    -.0761844    -.053466
                       |
                  race |
                    2  |  -.6069342   .1140528    -5.32   0.000    -.8304736   -.3833949
                    3  |  -.1947916   .0748669    -2.60   0.009    -.3415281   -.0480551
                    4  |  -.3734317   .0951604    -3.92   0.000    -.5599427   -.1869208
                    5  |  -.1074052   .1999645    -0.54   0.591    -.4993286    .2845181
                       |
                 round |
                    5  |   .1468429    .023539     6.24   0.000     .1007074    .1929784
                    6  |   -.082883    .024006    -3.45   0.001    -.1299339    -.035832
                       |
                 _cons |   3.492408   .0755697    46.21   0.000     3.344294    3.640522
-----------------------+----------------------------------------------------------------
               sigma_u |  .25298646
               sigma_e |  1.0924934
                   rho |  .05089456   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store condlive3

. xtreg condlive c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0443                                         min =        219
     Between = 0.2590                                         avg =    1,093.0
     Overall = 0.0533                                         max =      2,337

                                                Wald chi2(11)     =    1015.53
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
               condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0177182   .0060577     2.92   0.003     .0058453    .0295912
                educlvl |  -.0458649   .0073941    -6.20   0.000     -.060357   -.0313727
                        |
c.lnfdi_capex#c.educlvl |  -.0017971     .00125    -1.44   0.151    -.0042471    .0006529
                        |
                    age |   .0049643   .0005786     8.58   0.000     .0038303    .0060983
                 female |  -.0687682   .0153891    -4.47   0.000    -.0989303   -.0386061
                 unempl |  -.1508207   .0168164    -8.97   0.000    -.1837802   -.1178612
                  urban |  -.0515372   .0189825    -2.71   0.007    -.0887423   -.0143322
               interest |  -.0260132   .0074272    -3.50   0.000    -.0405701   -.0114562
                   news |  -.0663985   .0057813   -11.48   0.000    -.0777297   -.0550672
                        |
                  round |
                     5  |   .1459343    .023468     6.22   0.000     .0999378    .1919308
                     6  |  -.0877791   .0239958    -3.66   0.000    -.1348101   -.0407481
                        |
                  _cons |   3.426338   .0735668    46.57   0.000      3.28215    3.570526
------------------------+----------------------------------------------------------------
                sigma_u |  .25894182
                sigma_e |  1.0936393
                    rho |  .05308446   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store condlive4

. xtoprobit condlive c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -32442.445  
Iteration 1:   log likelihood = -31835.215  
Iteration 2:   log likelihood = -31835.077  
Iteration 3:   log likelihood = -31835.077  

Refining starting values:

Grid node 0:   log likelihood = -31390.028

Fitting full model:

Iteration 0:   log likelihood = -31390.028  (not concave)
Iteration 1:   log likelihood = -31384.254  (not concave)
Iteration 2:   log likelihood = -31378.461  (not concave)
Iteration 3:   log likelihood = -31373.039  (not concave)
Iteration 4:   log likelihood = -31368.064  (not concave)
Iteration 5:   log likelihood = -31365.994  (not concave)
Iteration 6:   log likelihood = -31364.476  
Iteration 7:   log likelihood = -31338.227  
Iteration 8:   log likelihood =  -31338.09  
Iteration 9:   log likelihood = -31338.089  

Random-effects ordered probit regression            Number of obs    =  21,859
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     219
                                                                 avg = 1,093.0
                                                                 max =   2,337

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  960.14
Log likelihood = -31338.089                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
              condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0167539   .0053313     3.14   0.002     .0063047     .027203
               educlvl |  -.0380796   .0074939    -5.08   0.000    -.0527674   -.0233919
                       |
c.lnfdi_jobs#c.educlvl |  -.0022978   .0011062    -2.08   0.038     -.004466   -.0001296
                       |
                   age |   .0047344   .0005601     8.45   0.000     .0036366    .0058322
                female |  -.0649059    .014866    -4.37   0.000    -.0940427   -.0357691
                unempl |  -.1426788   .0162526    -8.78   0.000    -.1745333   -.1108243
                 urban |  -.0469033   .0182525    -2.57   0.010    -.0826776   -.0111291
              interest |   -.025526   .0071885    -3.55   0.000    -.0396152   -.0114368
                  news |   -.063226   .0055885   -11.31   0.000    -.0741792   -.0522728
                       |
                 round |
                    5  |   .1346365   .0227277     5.92   0.000      .090091    .1791821
                    6  |  -.0890107   .0231585    -3.84   0.000    -.1344004   -.0436209
-----------------------+----------------------------------------------------------------
                 /cut1 |  -2.005588   .0689784                     -2.140783   -1.870393
                 /cut2 |  -.6495524   .0676301                     -.7821049   -.5169999
                 /cut3 |   .0105364   .0675287                     -.1218174    .1428902
                 /cut4 |   .8586738   .0677238                      .7259376    .9914101
-----------------------+----------------------------------------------------------------
             /sigma2_u |   .0506501   .0163571                      .0268962    .0953829
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 993.98       Prob >= chibar2 = 0.0000

. est store condlive5

. xtreg condlive c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0443                                         min =        219
     Between = 0.2728                                         avg =    1,093.0
     Overall = 0.0539                                         max =      2,337

                                                Wald chi2(11)     =    1018.56
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                 condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0156016   .0047473     3.29   0.001      .006297    .0249061
                  educlvl |  -.0409436   .0056647    -7.23   0.000    -.0520461    -.029841
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0033231   .0009621    -3.45   0.001    -.0052087   -.0014374
                          |
                      age |   .0050522   .0005787     8.73   0.000      .003918    .0061864
                   female |  -.0682979   .0153913    -4.44   0.000    -.0984643   -.0381315
                   unempl |  -.1501976   .0168186    -8.93   0.000    -.1831614   -.1172338
                    urban |   -.043577   .0193992    -2.25   0.025    -.0815986   -.0055553
                 interest |  -.0265358   .0074236    -3.57   0.000    -.0410859   -.0119858
                     news |  -.0658027   .0057714   -11.40   0.000    -.0771144    -.054491
                          |
                    round |
                       5  |   .1565586   .0232007     6.75   0.000     .1110861    .2020311
                       6  |  -.0700341   .0233703    -3.00   0.003     -.115839   -.0242292
                          |
                    _cons |   3.427664   .0603386    56.81   0.000     3.309402    3.545925
--------------------------+----------------------------------------------------------------
                  sigma_u |  .19567458
                  sigma_e |  1.0936369
                      rho |  .03101969   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store condlive6

. * Currently Bad Economic Situation
. xtreg condecon c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,889
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0191                                         min =        222
     Between = 0.0000                                         avg =    1,094.5
     Overall = 0.0155                                         max =      2,355

                                                Wald chi2(9)      =     426.60
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0240606   .0056659     4.25   0.000     .0129556    .0351655
               educlvl |    .005835   .0077994     0.75   0.454    -.0094515    .0211214
                       |
c.lnfdi_jobs#c.educlvl |  -.0020535   .0011754    -1.75   0.081    -.0043572    .0002502
                       |
                   age |   .0030975   .0005939     5.22   0.000     .0019335    .0042615
                female |  -.0172952   .0155862    -1.11   0.267    -.0478437    .0132532
                unempl |  -.0666233   .0172957    -3.85   0.000    -.1005223   -.0327243
                 urban |   .0444271    .019186     2.32   0.021     .0068232     .082031
                       |
                 round |
                    5  |    .277558   .0242867    11.43   0.000      .229957     .325159
                    6  |  -.0251509   .0247411    -1.02   0.309    -.0736425    .0233408
                       |
                 _cons |   3.139446   .1207989    25.99   0.000     2.902685    3.376208
-----------------------+----------------------------------------------------------------
               sigma_u |  .49842144
               sigma_e |  1.1263582
                   rho |  .16374849   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store condecon1

. xtreg condecon c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,667
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0209                                         min =        219
     Between = 0.0000                                         avg =    1,083.3
     Overall = 0.0175                                         max =      2,332

                                                Wald chi2(11)     =     461.41
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0257264   .0057337     4.49   0.000     .0144886    .0369643
               educlvl |   .0137694   .0080451     1.71   0.087    -.0019988    .0295376
                       |
c.lnfdi_jobs#c.educlvl |  -.0023577   .0011862    -1.99   0.047    -.0046826   -.0000328
                       |
                   age |   .0032796   .0005996     5.47   0.000     .0021043    .0044548
                female |   -.034709    .015903    -2.18   0.029    -.0658784   -.0035397
                unempl |  -.0711047   .0173989    -4.09   0.000    -.1052059   -.0370036
                 urban |   .0545116   .0195693     2.79   0.005     .0161565    .0928668
              interest |  -.0455075    .007686    -5.92   0.000    -.0605718   -.0304432
                  news |   -.015779   .0059705    -2.64   0.008     -.027481    -.004077
                       |
                 round |
                    5  |   .2706319   .0243901    11.10   0.000     .2228282    .3184357
                    6  |  -.0296407   .0248403    -1.19   0.233    -.0783268    .0190454
                       |
                 _cons |   3.202355   .1082679    29.58   0.000     2.990154    3.414556
-----------------------+----------------------------------------------------------------
               sigma_u |  .43366607
               sigma_e |  1.1251527
                   rho |  .12934095   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store condecon2

. xtreg condecon c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,571
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0212                                         min =        130
     Between = 0.0416                                         avg =    1,078.5
     Overall = 0.0187                                         max =      2,332

                                                Wald chi2(15)     =     468.11
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0253847   .0057396     4.42   0.000     .0141353    .0366341
               educlvl |   .0135653   .0080481     1.69   0.092    -.0022086    .0293393
                       |
c.lnfdi_jobs#c.educlvl |   -.002301    .001188    -1.94   0.053    -.0046295    .0000275
                       |
                   age |   .0033533   .0006012     5.58   0.000      .002175    .0045316
                female |  -.0327837   .0159359    -2.06   0.040    -.0640176   -.0015498
                unempl |  -.0699432   .0174379    -4.01   0.000    -.1041209   -.0357656
                 urban |   .0568215   .0196089     2.90   0.004     .0183888    .0952542
              interest |  -.0440402      .0077    -5.72   0.000     -.059132   -.0289485
                  news |   -.016147   .0059893    -2.70   0.007    -.0278858   -.0044083
                       |
                  race |
                    2  |   .0563289   .1196182     0.47   0.638    -.1781185    .2907764
                    3  |   .0466648   .0773259     0.60   0.546    -.1048912    .1982209
                    4  |  -.1517851   .1024441    -1.48   0.138    -.3525719    .0490017
                    5  |  -.3235265   .2059663    -1.57   0.116    -.7272131    .0801602
                       |
                 round |
                    5  |   .2717436   .0244036    11.14   0.000     .2239135    .3195736
                    6  |  -.0284084   .0248672    -1.14   0.253    -.0771472    .0203304
                       |
                 _cons |   3.221259   .0829083    38.85   0.000     3.058761    3.383756
-----------------------+----------------------------------------------------------------
               sigma_u |   .2874437
               sigma_e |  1.1245463
                   rho |  .06132879   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store condecon3

. xtreg condecon c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,667
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0211                                         min =        219
     Between = 0.0008                                         avg =    1,083.3
     Overall = 0.0174                                         max =      2,332

                                                Wald chi2(11)     =     466.60
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
               condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0270636   .0063043     4.29   0.000     .0147075    .0394197
                educlvl |   .0086478   .0076707     1.13   0.260    -.0063866    .0236822
                        |
c.lnfdi_capex#c.educlvl |  -.0017638    .001297    -1.36   0.174    -.0043058    .0007782
                        |
                    age |   .0032729   .0005996     5.46   0.000     .0020978     .004448
                 female |  -.0350433   .0159012    -2.20   0.028    -.0662091   -.0038775
                 unempl |    -.07199    .017397    -4.14   0.000    -.1060876   -.0378925
                  urban |   .0503708   .0196531     2.56   0.010     .0118514    .0888902
               interest |  -.0453494   .0076858    -5.90   0.000    -.0604133   -.0302855
                   news |  -.0159944   .0059707    -2.68   0.007    -.0276967   -.0042922
                        |
                  round |
                     5  |   .2707779   .0243118    11.14   0.000     .2231277    .3184282
                     6  |  -.0334011    .024841    -1.34   0.179    -.0820886    .0152865
                        |
                  _cons |   3.221926   .1065579    30.24   0.000     3.013076    3.430776
------------------------+----------------------------------------------------------------
                sigma_u |  .42742075
                sigma_e |  1.1250223
                    rho |  .12613442   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store condecon4

. xtoprobit condecon c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -32159.466  
Iteration 1:   log likelihood = -31912.682  
Iteration 2:   log likelihood = -31912.668  
Iteration 3:   log likelihood = -31912.668  

Refining starting values:

Grid node 0:   log likelihood =   -30515.4

Fitting full model:

Iteration 0:   log likelihood =   -30515.4  (not concave)
Iteration 1:   log likelihood = -30510.367  (not concave)
Iteration 2:   log likelihood = -30505.332  (not concave)
Iteration 3:   log likelihood = -30501.121  (not concave)
Iteration 4:   log likelihood = -30497.648  (not concave)
Iteration 5:   log likelihood = -30496.322  
Iteration 6:   log likelihood = -30371.037  
Iteration 7:   log likelihood = -30369.811  
Iteration 8:   log likelihood = -30369.693  
Iteration 9:   log likelihood = -30369.692  

Random-effects ordered probit regression            Number of obs    =  21,667
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     219
                                                                 avg = 1,083.3
                                                                 max =   2,332

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  455.16
Log likelihood = -30369.692                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
              condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0264143   .0054528     4.84   0.000     .0157271    .0371016
               educlvl |   .0159775   .0076442     2.09   0.037     .0009951    .0309599
                       |
c.lnfdi_jobs#c.educlvl |   -.002558   .0011306    -2.26   0.024    -.0047739   -.0003422
                       |
                   age |   .0030701   .0005714     5.37   0.000     .0019502      .00419
                female |  -.0329792   .0151214    -2.18   0.029    -.0626166   -.0033417
                unempl |  -.0653754   .0165293    -3.96   0.000    -.0977721   -.0329786
                 urban |   .0562376   .0186216     3.02   0.003     .0197399    .0927354
              interest |  -.0431722   .0073335    -5.89   0.000    -.0575456   -.0287989
                  news |  -.0148089   .0056827    -2.61   0.009    -.0259469    -.003671
                       |
                 round |
                    5  |   .2419917   .0232195    10.42   0.000     .1964823     .287501
                    6  |  -.0476169   .0235499    -2.02   0.043    -.0937738   -.0014599
-----------------------+----------------------------------------------------------------
                 /cut1 |  -1.675827    .102189                     -1.876114    -1.47554
                 /cut2 |  -.3523513   .1012625                      -.550822   -.1538805
                 /cut3 |   .1555041   .1012366                     -.0429159    .3539241
                 /cut4 |   .9776878   .1013842                      .7789784    1.176397
-----------------------+----------------------------------------------------------------
             /sigma2_u |   .1625732   .0517821                      .0870819    .3035079
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 3085.95      Prob >= chibar2 = 0.0000

. est store condecon5

. xtreg condecon c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,667
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0203                                         min =        219
     Between = 0.0013                                         avg =    1,083.3
     Overall = 0.0156                                         max =      2,332

                                                Wald chi2(11)     =     446.87
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                 condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0187435   .0049391     3.79   0.000      .009063    .0284239
                  educlvl |    .010051   .0058707     1.71   0.087    -.0014553    .0215573
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0024244    .000998    -2.43   0.015    -.0043803   -.0004684
                          |
                      age |   .0033396   .0005999     5.57   0.000     .0021637    .0045154
                   female |  -.0345801   .0159067    -2.17   0.030    -.0657566   -.0034035
                   unempl |  -.0712777   .0174053    -4.10   0.000    -.1053915   -.0371638
                    urban |   .0558573   .0201024     2.78   0.005     .0164574    .0952572
                 interest |  -.0464195   .0076856    -6.04   0.000     -.061483   -.0313561
                     news |  -.0147809   .0059669    -2.48   0.013    -.0264758    -.003086
                          |
                    round |
                       5  |   .2876625   .0240641    11.95   0.000     .2404977    .3348273
                       6  |  -.0061076   .0242192    -0.25   0.801    -.0535763    .0413612
                          |
                    _cons |   3.251016   .1069726    30.39   0.000     3.041354    3.460678
--------------------------+----------------------------------------------------------------
                  sigma_u |  .43782075
                  sigma_e |  1.1255207
                      rho |  .13142895   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store condecon6

. 
. *** TABLE A-4: Economic Grievances - Main Models
. esttab problem2 problem4 problem5 condlive2 condlive4 condlive5 condecon2 condecon4 condecon5, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl lnfdi_capex c.lnfdi_capex#c.educlvl age female unempl urba
> n interest news) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 lnfdi_capex "FDI capital expenditure" ///
>                 c.lnfdi_jobs#c.educlvl "FDI jobs * education" ///
>                 c.lnfdi_capex#c.educlvl "FDI capex * education" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                  problem         problem         problem        condlive        condlive        condlive        con
> decon        condecon        condecon   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
Education ~l        0.024***        0.023***        0.064***       -0.043***       -0.046***       -0.038***        
> 0.014*          0.009           0.016** 
                   (0.00)          (0.00)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
FDI jobs c~d        0.013***                        0.036***        0.016***                        0.017***        
> 0.026***                        0.026***
                   (0.00)                          (0.01)          (0.01)                          (0.01)          (
> 0.01)                          (0.01)   
FDI jobs *~n       -0.003***                       -0.008***       -0.002*                         -0.002**        -
> 0.002**                        -0.003** 
                   (0.00)                          (0.00)          (0.00)                          (0.00)          (
> 0.00)                          (0.00)   
FDI capita~e                        0.015***                                        0.018***                        
>                 0.027***                
                                   (0.00)                                          (0.01)                           
>                (0.01)                   
FDI capex ~n                       -0.003***                                       -0.002                           
>                -0.002                   
                                   (0.00)                                          (0.00)                           
>                (0.00)                   
Age in years       -0.002***       -0.002***       -0.004***        0.005***        0.005***        0.005***        
> 0.003***        0.003***        0.003***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female              0.003           0.003           0.008          -0.069***       -0.069***       -0.065***       -
> 0.035**        -0.035**        -0.033** 
                   (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (0.01)          (
> 0.02)          (0.02)          (0.02)   
Unemployed         -0.029***       -0.029***       -0.075***       -0.150***       -0.151***       -0.143***       -
> 0.071***       -0.072***       -0.065***
                   (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
Urban resi~t        0.037***        0.036***        0.096***       -0.048**        -0.052***       -0.047**         
> 0.055***        0.050**         0.056***
                   (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
Political ~t       -0.008**        -0.008**        -0.020**        -0.026***       -0.026***       -0.026***       -
> 0.046***       -0.045***       -0.043***
                   (0.00)          (0.00)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Newspaper ~n        0.005**         0.005**         0.013*         -0.066***       -0.066***       -0.063***       -
> 0.016***       -0.016***       -0.015***
                   (0.00)          (0.00)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
lnsig2u                                            -2.232***                                                        
>                                         
                                                   (0.32)                                                           
>                                         
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        21644           21644           21644           21859           21859           21859           
> 21667           21667           21667   
\# of coun~s           20              20              20              20              20              20           
>    20              20              20   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** TABLE A-5: Economic Grievances - Robustness Tests
. esttab problem1 problem3 problem6 condlive1 condlive3 condlive6 condecon1 condecon3 condecon6, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p ll, fmt(0 0 3 2) labels("\# respondents" "\# of countries" "Prob > Chi2" "Log likelihood")) 
> ///
>         order(educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl age female unempl urban interest news *race) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 lnfdi_capex "FDI capital expenditure" ///
>                 c.lnfdi_jobs#c.educlvl "FDI jobs * education" ///
>                 c.lnfdi_capex#c.educlvl "FDI capex * education" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption" 1.race "Black/African" ///
>                 2.race "White/European" 3.race "Colored/mixed race" ///
>                 4.race "Arab/Lebanese/North African" 5.race "Other")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                  problem         problem         problem        condlive        condlive        condlive        con
> decon        condecon        condecon   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Education ~l        0.024***        0.024***        0.014***       -0.065***       -0.043***       -0.041***        
> 0.006           0.014*          0.010*  
                   (0.00)          (0.00)          (0.00)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
FDI jobs c~d        0.013***        0.013***                        0.010*          0.015***                        
> 0.024***        0.025***                
                   (0.00)          (0.00)                          (0.01)          (0.01)                          (
> 0.01)          (0.01)                   
FDI jobs *~n       -0.003***       -0.003***                       -0.001          -0.002                          -
> 0.002*         -0.002*                  
                   (0.00)          (0.00)                          (0.00)          (0.00)                          (
> 0.00)          (0.00)                   
Age in years       -0.002***       -0.002***       -0.002***        0.005***        0.005***        0.005***        
> 0.003***        0.003***        0.003***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female              0.003           0.003           0.003          -0.039***       -0.067***       -0.068***       -
> 0.017          -0.033**        -0.035** 
                   (0.01)          (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
Unemployed         -0.030***       -0.029***       -0.029***       -0.141***       -0.149***       -0.150***       -
> 0.067***       -0.070***       -0.071***
                   (0.01)          (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
Urban resi~t        0.039***        0.036***        0.037***       -0.090***       -0.047**        -0.044**         
> 0.044**         0.057***        0.056***
                   (0.01)          (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
Political ~t                       -0.008**        -0.008**                        -0.027***       -0.027***        
>                -0.044***       -0.046***
                                   (0.00)          (0.00)                          (0.01)          (0.01)           
>                (0.01)          (0.01)   
Newspaper ~n                        0.005*          0.005**                        -0.065***       -0.066***        
>                -0.016***       -0.015** 
                                   (0.00)          (0.00)                          (0.01)          (0.01)           
>                (0.01)          (0.01)   
Black/Afri~n                        0.000                                           0.000                           
>                 0.000                   
                                      (.)                                             (.)                           
>                   (.)                   
White/Euro~n                        0.041                                          -0.607***                        
>                 0.056                   
                                   (0.05)                                          (0.11)                           
>                (0.12)                   
Colored/mi~e                       -0.018                                          -0.195***                        
>                 0.047                   
                                   (0.03)                                          (0.07)                           
>                (0.08)                   
Arab/Leban~n                        0.024                                          -0.373***                        
>                -0.152                   
                                   (0.04)                                          (0.10)                           
>                (0.10)                   
Other                              -0.003                                          -0.107                           
>                -0.324                   
                                   (0.09)                                          (0.20)                           
>                (0.21)                   
lnfdi_job~10                                        0.009***                                        0.016***        
>                                 0.019***
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
c.lnfdi_jo~l                                       -0.002***                                       -0.003***        
>                                -0.002** 
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        21876           21550           21644           22094           21763           21859           
> 21889           21571           21667   
\# of coun~s           20              20              20              20              20              20           
>    20              20              20   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
Log likeli~d                                                                                                        
>                                         
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** Set 2: Distrust in State Institutions
. * Distrust in Electoral Commission
. xtreg trustelec c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,738
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0133                                         min =        200
     Between = 0.1526                                         avg =      986.2
     Overall = 0.0164                                         max =      2,281

                                                Wald chi2(9)      =     254.25
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0098363   .0027094     3.63   0.000      .004526    .0151466
               educlvl |   .0191184   .0035608     5.37   0.000     .0121394    .0260975
                       |
c.lnfdi_jobs#c.educlvl |  -.0017651   .0005649    -3.12   0.002    -.0028722    -.000658
                       |
                   age |  -.0019376   .0002749    -7.05   0.000    -.0024764   -.0013987
                female |  -.0009238   .0071342    -0.13   0.897    -.0149066     .013059
                unempl |  -.0061471   .0080162    -0.77   0.443    -.0218585    .0095643
                 urban |    .056084   .0088297     6.35   0.000     .0387781    .0733899
                       |
                 round |
                    5  |    .033255   .0108821     3.06   0.002     .0119264    .0545835
                    6  |  -.0328314   .0109204    -3.01   0.003     -.054235   -.0114277
                       |
                 _cons |   .5038641   .0294821    17.09   0.000     .4460802     .561648
-----------------------+----------------------------------------------------------------
               sigma_u |  .08919731
               sigma_e |  .47811949
                   rho |  .03363353   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustelec1

. xtreg trustelec c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,583
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0175                                         min =        197
     Between = 0.1990                                         avg =      978.1
     Overall = 0.0217                                         max =      2,258

                                                Wald chi2(11)     =     336.39
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0097377   .0027323     3.56   0.000     .0043825    .0150929
               educlvl |   .0189582   .0036644     5.17   0.000     .0117762    .0261403
                       |
c.lnfdi_jobs#c.educlvl |  -.0017944   .0005684    -3.16   0.002    -.0029084   -.0006804
                       |
                   age |  -.0017365   .0002768    -6.27   0.000     -.002279    -.001194
                female |  -.0082069   .0072625    -1.13   0.258    -.0224411    .0060272
                unempl |  -.0066345   .0080427    -0.82   0.409    -.0223979    .0091289
                 urban |   .0522606   .0090092     5.80   0.000     .0346029    .0699182
              interest |  -.0307532   .0034905    -8.81   0.000    -.0375944    -.023912
                  news |   .0071711   .0027387     2.62   0.009     .0018033    .0125389
                       |
                 round |
                    5  |   .0319239   .0109021     2.93   0.003     .0105561    .0532916
                    6  |  -.0361977   .0109388    -3.31   0.001    -.0576374    -.014758
                       |
                 _cons |   .5458911   .0279087    19.56   0.000     .4911911     .600591
-----------------------+----------------------------------------------------------------
               sigma_u |  .07553513
               sigma_e |  .47701053
                   rho |  .02446168   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustelec2

. xtreg trustelec c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,490
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0178                                         min =        110
     Between = 0.1822                                         avg =      973.2
     Overall = 0.0217                                         max =      2,258

                                                Wald chi2(15)     =     338.40
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .010001   .0027385     3.65   0.000     .0046337    .0153684
               educlvl |   .0189769   .0036667     5.18   0.000     .0117902    .0261636
                       |
c.lnfdi_jobs#c.educlvl |  -.0018478   .0005699    -3.24   0.001    -.0029648   -.0007308
                       |
                   age |  -.0017176   .0002778    -6.18   0.000    -.0022621   -.0011731
                female |  -.0075866   .0072799    -1.04   0.297     -.021855    .0066817
                unempl |  -.0069827   .0080653    -0.87   0.387    -.0227904    .0088251
                 urban |   .0521447   .0090466     5.76   0.000     .0344136    .0698757
              interest |  -.0307853   .0034976    -8.80   0.000    -.0376405   -.0239302
                  news |   .0075401   .0027533     2.74   0.006     .0021437    .0129365
                       |
                  race |
                    2  |    .033611   .0615317     0.55   0.585     -.086989    .1542109
                    3  |  -.0436366   .0361133    -1.21   0.227    -.1144174    .0271442
                    4  |  -.0217043   .0520396    -0.42   0.677       -.1237    .0802915
                    5  |   .1268385   .0888565     1.43   0.153     -.047317     .300994
                       |
                 round |
                    5  |   .0315612   .0109302     2.89   0.004     .0101384     .052984
                    6  |  -.0375206   .0109701    -3.42   0.001    -.0590216   -.0160196
                       |
                 _cons |   .5486003   .0314489    17.44   0.000     .4869615    .6102391
-----------------------+----------------------------------------------------------------
               sigma_u |  .09149298
               sigma_e |  .47714944
                   rho |   .0354638   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustelec3

. xtreg trustelec c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,583
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0179                                         min =        197
     Between = 0.1942                                         avg =      978.1
     Overall = 0.0218                                         max =      2,258

                                                Wald chi2(11)     =     342.24
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
              trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0129551   .0029874     4.34   0.000     .0070998    .0188103
                educlvl |   .0198858   .0034898     5.70   0.000     .0130458    .0267257
                        |
c.lnfdi_capex#c.educlvl |  -.0022927   .0006129    -3.74   0.000    -.0034941   -.0010914
                        |
                    age |  -.0017366   .0002768    -6.27   0.000     -.002279   -.0011941
                 female |  -.0082614    .007261    -1.14   0.255    -.0224926    .0059698
                 unempl |  -.0067488   .0080414    -0.84   0.401    -.0225096    .0090121
                  urban |   .0500508   .0090479     5.53   0.000     .0323173    .0677843
               interest |   -.030622   .0034901    -8.77   0.000    -.0374625   -.0237815
                   news |   .0069971   .0027391     2.55   0.011     .0016284    .0123657
                        |
                  round |
                     5  |   .0309973   .0108696     2.85   0.004     .0096934    .0523013
                     6  |  -.0377223   .0109424    -3.45   0.001    -.0591691   -.0162755
                        |
                  _cons |   .5417097   .0279322    19.39   0.000     .4869636    .5964558
------------------------+----------------------------------------------------------------
                sigma_u |  .07822502
                sigma_e |  .47693185
                    rho |   .0261969   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store trustelec4

. xtprobit trustelec c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -12743.597  
Iteration 1:   log likelihood =  -12519.49  
Iteration 2:   log likelihood = -12519.425  
Iteration 3:   log likelihood = -12519.425  

Fitting full model:

rho =  0.0     log likelihood = -12519.425
rho =  0.1     log likelihood = -12083.922
rho =  0.2     log likelihood = -12118.673

Iteration 0:   log likelihood = -12066.828  
Iteration 1:   log likelihood = -12046.868  
Iteration 2:   log likelihood = -12046.852  
Iteration 3:   log likelihood = -12046.852  

Random-effects probit regression                     Number of obs    = 18,583
Group variable: cname                                Number of groups =     19

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    197
                                                                  avg =  978.1
                                                                  max =  2,258

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(11)    = 322.70
Log likelihood = -12046.852                          Prob > chi2      = 0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0258788   .0073287     3.53   0.000     .0115148    .0402427
               educlvl |   .0496119   .0098623     5.03   0.000     .0302821    .0689417
                       |
c.lnfdi_jobs#c.educlvl |  -.0047635   .0015313    -3.11   0.002    -.0077648   -.0017622
                       |
                   age |  -.0046147   .0007464    -6.18   0.000    -.0060777   -.0031517
                female |  -.0218504   .0196296    -1.11   0.266    -.0603237    .0166228
                unempl |  -.0184464   .0217351    -0.85   0.396    -.0610464    .0241535
                 urban |   .1386468   .0242331     5.72   0.000     .0911507    .1861429
              interest |  -.0825382   .0094489    -8.74   0.000    -.1010576   -.0640187
                  news |   .0199036   .0074229     2.68   0.007     .0053549    .0344524
                       |
                 round |
                    5  |   .0837043   .0295771     2.83   0.005     .0257342    .1416744
                    6  |  -.0976622   .0296277    -3.30   0.001    -.1557314    -.039593
                       |
                 _cons |   .1292632   .0896826     1.44   0.149    -.0465115     .305038
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.444932   .3337316                     -3.099034    -1.79083
-----------------------+----------------------------------------------------------------
               sigma_u |   .2945031   .0491425                      .2123506    .4084381
                   rho |     .07981   .0245094                      .0431471     .142971
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 945.15                 Prob >= chibar2 = 0.000

. est store trustelec5

. xtreg trustelec c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,583
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0174                                         min =        197
     Between = 0.1955                                         avg =      978.1
     Overall = 0.0216                                         max =      2,258

                                                Wald chi2(11)     =     333.21
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0079233   .0024407     3.25   0.001     .0031396     .012707
                  educlvl |    .014218   .0026958     5.27   0.000     .0089343    .0195018
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0013662   .0004907    -2.78   0.005    -.0023279   -.0004045
                          |
                      age |  -.0017176   .0002769    -6.20   0.000    -.0022603    -.001175
                   female |  -.0084818    .007262    -1.17   0.243     -.022715    .0057515
                   unempl |  -.0065806   .0080436    -0.82   0.413    -.0223459    .0091846
                    urban |    .050392   .0092929     5.42   0.000     .0321783    .0686058
                 interest |  -.0308279     .00349    -8.83   0.000    -.0376683   -.0239876
                     news |   .0073291   .0027374     2.68   0.007     .0019639    .0126943
                          |
                    round |
                       5  |   .0342101   .0107539     3.18   0.001      .013133    .0552873
                       6  |  -.0335947   .0106519    -3.15   0.002    -.0544721   -.0127173
                          |
                    _cons |   .5681027   .0274875    20.67   0.000     .5142281    .6219773
--------------------------+----------------------------------------------------------------
                  sigma_u |  .08400905
                  sigma_e |  .47704108
                      rho |  .03007991   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store trustelec6

. * Distrust in Leader
. xtreg trustlead c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,063
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0188                                         min =        214
     Between = 0.0003                                         avg =    1,053.2
     Overall = 0.0206                                         max =      2,316

                                                Wald chi2(9)      =     403.25
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0066226   .0024041     2.75   0.006     .0019107    .0113345
               educlvl |    .016643   .0033323     4.99   0.000     .0101117    .0231742
                       |
c.lnfdi_jobs#c.educlvl |  -.0008309   .0005003    -1.66   0.097    -.0018115    .0001496
                       |
                   age |   -.002221   .0002551    -8.71   0.000     -.002721    -.001721
                female |   .0014458   .0066997     0.22   0.829    -.0116852    .0145769
                unempl |  -.0008678   .0074171    -0.12   0.907    -.0154049    .0136694
                 urban |    .055719   .0082815     6.73   0.000     .0394875    .0719505
                       |
                 round |
                    5  |   .0545157   .0103517     5.27   0.000     .0342266    .0748047
                    6  |   -.033086   .0105524    -3.14   0.002    -.0537684   -.0124036
                       |
                 _cons |   .3891645   .0323814    12.02   0.000     .3256981    .4526308
-----------------------+----------------------------------------------------------------
               sigma_u |  .11432508
               sigma_e |  .47476667
                   rho |  .05480785   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustlead1

. xtreg trustlead c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0214                                         min =        211
     Between = 0.0091                                         avg =    1,043.0
     Overall = 0.0203                                         max =      2,294

                                                Wald chi2(11)     =     453.38
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0063255    .002429     2.60   0.009     .0015646    .0110863
               educlvl |   .0153434   .0034325     4.47   0.000     .0086159    .0220709
                       |
c.lnfdi_jobs#c.educlvl |   -.000845   .0005041    -1.68   0.094    -.0018331     .000143
                       |
                   age |  -.0020464   .0002573    -7.95   0.000    -.0025508   -.0015421
                female |  -.0011829   .0068309    -0.17   0.863    -.0145712    .0122054
                unempl |  -.0008925   .0074545    -0.12   0.905    -.0155031    .0137181
                 urban |   .0507245   .0084438     6.01   0.000      .034175    .0672741
              interest |  -.0229246   .0033128    -6.92   0.000    -.0294175   -.0164317
                  news |   .0107462   .0025639     4.19   0.000     .0057211    .0157713
                       |
                 round |
                    5  |   .0509599   .0103856     4.91   0.000     .0306046    .0713153
                    6  |  -.0364752   .0105845    -3.45   0.001    -.0572204   -.0157301
                       |
                 _cons |   .4190262   .0297561    14.08   0.000     .3607054     .477347
-----------------------+----------------------------------------------------------------
               sigma_u |  .09648029
               sigma_e |  .47400348
                   rho |  .03978177   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustlead2

. xtreg trustlead c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,764
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0218                                         min =        122
     Between = 0.0043                                         avg =    1,038.2
     Overall = 0.0253                                         max =      2,294

                                                Wald chi2(15)     =     460.97
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0062693   .0024319     2.58   0.010     .0015029    .0110357
               educlvl |   .0154309   .0034333     4.49   0.000     .0087017    .0221601
                       |
c.lnfdi_jobs#c.educlvl |  -.0008662   .0005049    -1.72   0.086    -.0018557    .0001233
                       |
                   age |  -.0020181    .000258    -7.82   0.000    -.0025237   -.0015125
                female |  -.0005894    .006845    -0.09   0.931    -.0140054    .0128265
                unempl |  -.0010783   .0074722    -0.14   0.885    -.0157236    .0135669
                 urban |   .0523909   .0084647     6.19   0.000     .0358003    .0689814
              interest |  -.0228687   .0033185    -6.89   0.000    -.0293728   -.0163645
                  news |   .0105721   .0025731     4.11   0.000     .0055289    .0156152
                       |
                  race |
                    2  |   .0495717   .0514993     0.96   0.336     -.051365    .1505084
                    3  |  -.0423341   .0329977    -1.28   0.200    -.1070083    .0223401
                    4  |   .0554266   .0389071     1.42   0.154    -.0208299    .1316832
                    5  |  -.1793636   .0882581    -2.03   0.042    -.3523464   -.0063809
                       |
                 round |
                    5  |   .0508745   .0103937     4.89   0.000     .0305032    .0712458
                    6  |  -.0377752   .0106002    -3.56   0.000    -.0585512   -.0169992
                       |
                 _cons |   .4063151   .0300398    13.53   0.000     .3474382    .4651921
-----------------------+----------------------------------------------------------------
               sigma_u |  .09202928
               sigma_e |  .47394917
                   rho |  .03633415   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustlead3

. xtreg trustlead c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0213                                         min =        211
     Between = 0.0106                                         avg =    1,043.0
     Overall = 0.0198                                         max =      2,294

                                                Wald chi2(11)     =     450.06
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
              trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0060784    .002668     2.28   0.023     .0008493    .0113076
                educlvl |     .01617   .0032694     4.95   0.000     .0097622    .0225778
                        |
c.lnfdi_capex#c.educlvl |  -.0011367   .0005499    -2.07   0.039    -.0022145   -.0000588
                        |
                    age |  -.0020336   .0002574    -7.90   0.000     -.002538   -.0015292
                 female |  -.0011059   .0068328    -0.16   0.871     -.014498    .0122862
                 unempl |  -.0008185   .0074562    -0.11   0.913    -.0154324    .0137955
                  urban |   .0526173   .0084766     6.21   0.000     .0360034    .0692311
               interest |  -.0229889   .0033134    -6.94   0.000     -.029483   -.0164948
                   news |   .0107836   .0025637     4.21   0.000      .005759    .0158083
                        |
                  round |
                     5  |    .053304   .0103507     5.15   0.000      .033017    .0735909
                     6  |  -.0329823   .0105808    -3.12   0.002    -.0537203   -.0122443
                        |
                  _cons |   .4204324   .0273731    15.36   0.000     .3667821    .4740827
------------------------+----------------------------------------------------------------
                sigma_u |   .0833257
                sigma_e |  .47403155
                    rho |  .02997281   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store trustlead4

. xtprobit trustlead c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -14343.917  
Iteration 1:   log likelihood = -14087.024  
Iteration 2:   log likelihood = -14086.896  
Iteration 3:   log likelihood = -14086.896  

Fitting full model:

rho =  0.0     log likelihood = -14086.896
rho =  0.1     log likelihood = -13456.036
rho =  0.2     log likelihood = -13512.293

Iteration 0:   log likelihood = -13427.411  
Iteration 1:   log likelihood = -13388.745  
Iteration 2:   log likelihood = -13388.743  
Iteration 3:   log likelihood = -13388.743  

Random-effects probit regression                    Number of obs    =  20,859
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     211
                                                                 avg = 1,043.0
                                                                 max =   2,294

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  440.00
Log likelihood = -13388.743                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
             trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0174085   .0066739     2.61   0.009      .004328     .030489
               educlvl |   .0401229   .0093574     4.29   0.000     .0217827    .0584631
                       |
c.lnfdi_jobs#c.educlvl |  -.0022394   .0013748    -1.63   0.103    -.0049339    .0004551
                       |
                   age |  -.0057456    .000708    -8.12   0.000    -.0071332    -.004358
                female |  -.0041467   .0186594    -0.22   0.824    -.0407184     .032425
                unempl |   -.003801   .0203292    -0.19   0.852    -.0436455    .0360436
                 urban |   .1403341   .0231335     6.07   0.000     .0949932     .185675
              interest |  -.0634645   .0090588    -7.01   0.000    -.0812195   -.0457096
                  news |   .0306141   .0070066     4.37   0.000     .0168815    .0443467
                       |
                 round |
                    5  |   .1372572   .0285601     4.81   0.000     .0812805     .193234
                    6  |  -.0952956   .0291736    -3.27   0.001    -.1524749   -.0381163
                       |
                 _cons |  -.2118935   .0928641    -2.28   0.023    -.3939038   -.0298833
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.214996   .3216899                     -2.845497   -1.584495
-----------------------+----------------------------------------------------------------
               sigma_u |   .3303845   .0531407                      .2410506    .4528258
                   rho |   .0984119   .0285426                      .0549146    .1701598
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 1396.31                Prob >= chibar2 = 0.000

. est store trustlead5

. xtreg trustlead c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0218                                         min =        211
     Between = 0.0099                                         avg =    1,043.0
     Overall = 0.0203                                         max =      2,294

                                                Wald chi2(11)     =     461.20
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0066384   .0021052     3.15   0.002     .0025123    .0107644
                  educlvl |   .0129892   .0025108     5.17   0.000     .0080683    .0179102
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0006606   .0004258    -1.55   0.121    -.0014952    .0001741
                          |
                      age |  -.0020511   .0002573    -7.97   0.000    -.0025553   -.0015468
                   female |  -.0014203   .0068284    -0.21   0.835    -.0148038    .0119632
                   unempl |  -.0010346   .0074531    -0.14   0.890    -.0156424    .0135732
                    urban |   .0453409   .0086567     5.24   0.000     .0283741    .0623076
                 interest |  -.0230737   .0033108    -6.97   0.000    -.0295628   -.0165847
                     news |    .010775   .0025611     4.21   0.000     .0057554    .0157945
                          |
                    round |
                       5  |   .0523534   .0102465     5.11   0.000     .0322706    .0724361
                       6  |  -.0348184   .0103085    -3.38   0.001    -.0550227   -.0146142
                          |
                    _cons |   .4321879   .0297673    14.52   0.000      .373845    .4905308
--------------------------+----------------------------------------------------------------
                  sigma_u |  .10471022
                  sigma_e |  .47391855
                      rho |  .04654478   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store trustlead6

. * Distrust in Ruling Party
. xtreg trustpart c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,705
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0209                                         min =        209
     Between = 0.1029                                         avg =    1,037.1
     Overall = 0.0266                                         max =      2,311

                                                Wald chi2(9)      =     422.69
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0103369    .002488     4.15   0.000     .0054604    .0152134
               educlvl |   .0223457   .0034561     6.47   0.000     .0155719    .0291196
                       |
c.lnfdi_jobs#c.educlvl |  -.0012023   .0005178    -2.32   0.020    -.0022172   -.0001874
                       |
                   age |   -.002064   .0002594    -7.96   0.000    -.0025724   -.0015556
                female |  -.0080271   .0067722    -1.19   0.236    -.0213004    .0052462
                unempl |  -.0146133   .0075198    -1.94   0.052    -.0293518    .0001251
                 urban |   .0558593   .0084442     6.62   0.000     .0393089    .0724097
                       |
                 round |
                    5  |   .0210563   .0102444     2.06   0.040     .0009776    .0411349
                    6  |   -.049851   .0105181    -4.74   0.000     -.070466    -.029236
                       |
                 _cons |    .552674   .0321752    17.18   0.000     .4896118    .6157363
-----------------------+----------------------------------------------------------------
               sigma_u |  .10878264
               sigma_e |  .46449399
                   rho |  .05199591   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustpart1

. xtreg trustpart c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,536
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0269                                         min =        206
     Between = 0.0496                                         avg =    1,028.2
     Overall = 0.0295                                         max =      2,290

                                                Wald chi2(11)     =     541.37
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0091813    .002503     3.67   0.000     .0042755    .0140872
               educlvl |   .0201943   .0035446     5.70   0.000      .013247    .0271416
                       |
c.lnfdi_jobs#c.educlvl |  -.0011215   .0005197    -2.16   0.031    -.0021401   -.0001028
                       |
                   age |   -.001848    .000261    -7.08   0.000    -.0023595   -.0013365
                female |  -.0122745   .0068856    -1.78   0.075      -.02577    .0012211
                unempl |   -.014705   .0075401    -1.95   0.051    -.0294834    .0000734
                 urban |   .0452122   .0086042     5.25   0.000     .0283482    .0620762
              interest |  -.0330453    .003335    -9.91   0.000    -.0395817   -.0265088
                  news |   .0155354   .0025755     6.03   0.000     .0104875    .0205833
                       |
                 round |
                    5  |   .0197332   .0102578     1.92   0.054    -.0003716    .0398381
                    6  |  -.0519232   .0105303    -4.93   0.000    -.0725622   -.0312842
                       |
                 _cons |   .6002221   .0303931    19.75   0.000     .5406527    .6597914
-----------------------+----------------------------------------------------------------
               sigma_u |  .09623867
               sigma_e |   .4629095
                   rho |  .04143146   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustpart2

. xtreg trustpart c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,444
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0261                                         min =        119
     Between = 0.1819                                         avg =    1,023.4
     Overall = 0.0399                                         max =      2,290

                                                Wald chi2(15)     =     629.81
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0068948   .0025334     2.72   0.006     .0019294    .0118601
               educlvl |   .0249662   .0035856     6.96   0.000     .0179387    .0319938
                       |
c.lnfdi_jobs#c.educlvl |  -.0011184   .0005284    -2.12   0.034     -.002154   -.0000828
                       |
                   age |  -.0018122   .0002647    -6.85   0.000    -.0023311   -.0012934
                female |  -.0158928   .0070094    -2.27   0.023    -.0296309   -.0021547
                unempl |  -.0037844   .0076558    -0.49   0.621    -.0187895    .0112207
                 urban |   .0618662    .008564     7.22   0.000     .0450811    .0786513
              interest |  -.0363045   .0033643   -10.79   0.000    -.0428985   -.0297105
                  news |   .0085544   .0025683     3.33   0.001     .0035207    .0135881
                       |
                  race |
                    2  |  -.0538287   .0508311    -1.06   0.290    -.1534559    .0457985
                    3  |  -.0462977   .0309081    -1.50   0.134    -.1068764    .0142811
                    4  |   .0848432   .0129002     6.58   0.000     .0595594    .1101271
                    5  |  -.0312332   .0861843    -0.36   0.717    -.2001513    .1376849
                       |
                 round |
                    5  |   .0334678   .0101829     3.29   0.001     .0135096     .053426
                    6  |  -.0404444   .0104114    -3.88   0.000    -.0608504   -.0200385
                       |
                 _cons |   .5481133   .0210049    26.09   0.000     .5069446    .5892821
-----------------------+----------------------------------------------------------------
               sigma_u |  .01552617
               sigma_e |  .46295671
                   rho |  .00112347   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store trustpart3

. xtreg trustpart c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,536
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0269                                         min =        206
     Between = 0.0561                                         avg =    1,028.2
     Overall = 0.0295                                         max =      2,290

                                                Wald chi2(11)     =     539.60
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
              trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0096066   .0027482     3.50   0.000     .0042202     .014993
                educlvl |   .0194782   .0033514     5.81   0.000     .0129095    .0260469
                        |
c.lnfdi_capex#c.educlvl |  -.0011866   .0005662    -2.10   0.036    -.0022963   -.0000769
                        |
                    age |  -.0018451    .000261    -7.07   0.000    -.0023566   -.0013336
                 female |  -.0122838   .0068855    -1.78   0.074    -.0257792    .0012116
                 unempl |  -.0149902   .0075403    -1.99   0.047    -.0297689   -.0002116
                  urban |   .0449263   .0086453     5.20   0.000     .0279817    .0618709
               interest |  -.0330254   .0033353    -9.90   0.000    -.0395624   -.0264883
                   news |   .0155938   .0025758     6.05   0.000     .0105452    .0206424
                        |
                  round |
                     5  |   .0206799   .0102249     2.02   0.043     .0006396    .0407203
                     6  |  -.0513149   .0105334    -4.87   0.000    -.0719599   -.0306699
                        |
                  _cons |   .6053638   .0306247    19.77   0.000     .5453405     .665387
------------------------+----------------------------------------------------------------
                sigma_u |  .09973966
                sigma_e |  .46293152
                    rho |  .04436049   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store trustpart4

. xtprobit trustpart c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -13210.329  
Iteration 1:   log likelihood = -12861.322  
Iteration 2:   log likelihood = -12861.056  
Iteration 3:   log likelihood = -12861.056  

Fitting full model:

rho =  0.0     log likelihood = -12861.056
rho =  0.1     log likelihood = -12158.605
rho =  0.2     log likelihood = -12226.525

Iteration 0:   log likelihood =  -12147.66  
Iteration 1:   log likelihood = -12099.434  
Iteration 2:   log likelihood =   -12099.4  
Iteration 3:   log likelihood =   -12099.4  

Random-effects probit regression                    Number of obs    =  19,536
Group variable: cname                               Number of groups =      19

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     206
                                                                 avg = 1,028.2
                                                                 max =   2,290

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  513.96
Log likelihood = -12099.4                           Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
             trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0251153   .0070779     3.55   0.000     .0112428    .0389877
               educlvl |   .0548366   .0101037     5.43   0.000     .0350338    .0746395
                       |
c.lnfdi_jobs#c.educlvl |  -.0028134   .0014888    -1.89   0.059    -.0057313    .0001046
                       |
                   age |  -.0050685   .0007382    -6.87   0.000    -.0065155   -.0036216
                female |  -.0343838   .0195452    -1.76   0.079    -.0726917    .0039241
                unempl |  -.0414162   .0214002    -1.94   0.053    -.0833598    .0005273
                 urban |   .1230136   .0242098     5.08   0.000     .0755632     .170464
              interest |  -.0935668   .0094856    -9.86   0.000    -.1121582   -.0749754
                  news |   .0436232   .0073319     5.95   0.000      .029253    .0579935
                       |
                 round |
                    5  |    .052063   .0290851     1.79   0.073    -.0049428    .1090687
                    6  |  -.1470489   .0297802    -4.94   0.000     -.205417   -.0886808
                       |
                 _cons |   .2812132   .1038769     2.71   0.007     .0776183    .4848081
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -1.975658   .3294007                     -2.621272   -1.330045
-----------------------+----------------------------------------------------------------
               sigma_u |   .3723842   .0613318                      .2696485     .514262
                   rho |   .1217824   .0352299                      .0677819     .209152
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 1523.31                Prob >= chibar2 = 0.000

. est store trustpart5

. xtreg trustpart c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,536
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0265                                         min =        206
     Between = 0.0558                                         avg =    1,028.2
     Overall = 0.0291                                         max =      2,290

                                                Wald chi2(11)     =     532.83
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |    .005917   .0021532     2.75   0.006     .0016968    .0101372
                  educlvl |   .0170556   .0025829     6.60   0.000     .0119932    .0221179
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0008265   .0004351    -1.90   0.057    -.0016792    .0000262
                          |
                      age |  -.0018326   .0002611    -7.02   0.000    -.0023443   -.0013208
                   female |  -.0122481   .0068859    -1.78   0.075    -.0257443    .0012481
                   unempl |  -.0148865   .0075419    -1.97   0.048    -.0296684   -.0001046
                    urban |   .0459974   .0088281     5.21   0.000     .0286947    .0633001
                 interest |  -.0332737   .0033349    -9.98   0.000    -.0398099   -.0267375
                     news |   .0159146    .002573     6.19   0.000     .0108715    .0209577
                          |
                    round |
                       5  |   .0248376    .010112     2.46   0.014     .0050185    .0446567
                       6  |  -.0450659   .0102543    -4.39   0.000     -.065164   -.0249679
                          |
                    _cons |   .6218866   .0293443    21.19   0.000     .5643728    .6794004
--------------------------+----------------------------------------------------------------
                  sigma_u |  .09884268
                  sigma_e |  .46301007
                      rho |  .04358668   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store trustpart6

. 
. *** TABLE A-6: Distrust in State Institutions - Main Models
. esttab trustelec2 trustelec4 trustelec5 trustlead2 trustlead4 trustlead5 trustpart2 trustpart4 trustpart5, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl lnfdi_capex c.lnfdi_capex#c.educlvl age female unempl urba
> n interest news) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 lnfdi_capex "FDI capital expenditure" ///
>                 c.lnfdi_jobs#c.educlvl "FDI jobs * education" ///
>                 c.lnfdi_capex#c.educlvl "FDI capex * education" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                trustelec       trustelec       trustelec       trustlead       trustlead       trustlead       trus
> tpart       trustpart       trustpart   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
Education ~l        0.019***        0.020***        0.050***        0.015***        0.016***        0.040***        
> 0.020***        0.019***        0.055***
                   (0.00)          (0.00)          (0.01)          (0.00)          (0.00)          (0.01)          (
> 0.00)          (0.00)          (0.01)   
FDI jobs c~d        0.010***                        0.026***        0.006***                        0.017***        
> 0.009***                        0.025***
                   (0.00)                          (0.01)          (0.00)                          (0.01)          (
> 0.00)                          (0.01)   
FDI jobs *~n       -0.002***                       -0.005***       -0.001*                         -0.002          -
> 0.001**                        -0.003*  
                   (0.00)                          (0.00)          (0.00)                          (0.00)          (
> 0.00)                          (0.00)   
FDI capita~e                        0.013***                                        0.006**                         
>                 0.010***                
                                   (0.00)                                          (0.00)                           
>                (0.00)                   
FDI capex ~n                       -0.002***                                       -0.001**                         
>                -0.001**                 
                                   (0.00)                                          (0.00)                           
>                (0.00)                   
Age in years       -0.002***       -0.002***       -0.005***       -0.002***       -0.002***       -0.006***       -
> 0.002***       -0.002***       -0.005***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female             -0.008          -0.008          -0.022          -0.001          -0.001          -0.004          -
> 0.012*         -0.012*         -0.034*  
                   (0.01)          (0.01)          (0.02)          (0.01)          (0.01)          (0.02)          (
> 0.01)          (0.01)          (0.02)   
Unemployed         -0.007          -0.007          -0.018          -0.001          -0.001          -0.004          -
> 0.015*         -0.015**        -0.041*  
                   (0.01)          (0.01)          (0.02)          (0.01)          (0.01)          (0.02)          (
> 0.01)          (0.01)          (0.02)   
Urban resi~t        0.052***        0.050***        0.139***        0.051***        0.053***        0.140***        
> 0.045***        0.045***        0.123***
                   (0.01)          (0.01)          (0.02)          (0.01)          (0.01)          (0.02)          (
> 0.01)          (0.01)          (0.02)   
Political ~t       -0.031***       -0.031***       -0.083***       -0.023***       -0.023***       -0.063***       -
> 0.033***       -0.033***       -0.094***
                   (0.00)          (0.00)          (0.01)          (0.00)          (0.00)          (0.01)          (
> 0.00)          (0.00)          (0.01)   
Newspaper ~n        0.007***        0.007**         0.020***        0.011***        0.011***        0.031***        
> 0.016***        0.016***        0.044***
                   (0.00)          (0.00)          (0.01)          (0.00)          (0.00)          (0.01)          (
> 0.00)          (0.00)          (0.01)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
lnsig2u                                            -2.445***                                       -2.215***        
>                                -1.976***
                                                   (0.33)                                          (0.32)           
>                                (0.33)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        18583           18583           18583           20859           20859           20859           
> 19536           19536           19536   
\# of coun~s           19              19              19              20              20              20           
>    19              19              19   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** TABLE A-7: Distrust in State Institutions - Robustness Tests
. esttab trustelec1 trustelec3 trustelec6 trustlead1 trustlead3 trustlead6 trustpart1 trustpart3 trustpart6, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p ll, fmt(0 0 3 2) labels("\# respondents" "\# of countries" "Prob > Chi2" "Log likelihood")) 
> ///
>         order(educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl age female unempl urban interest news *race) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 lnfdi_capex "FDI capital expenditure" ///
>                 c.lnfdi_jobs#c.educlvl "FDI jobs * education" ///
>                 c.lnfdi_capex#c.educlvl "FDI capex * education" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption" 1.race "Black/African" ///
>                 2.race "White/European" 3.race "Colored/mixed race" ///
>                 4.race "Arab/Lebanese/North African" 5.race "Other")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                trustelec       trustelec       trustelec       trustlead       trustlead       trustlead       trus
> tpart       trustpart       trustpart   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Education ~l        0.019***        0.019***        0.014***        0.017***        0.015***        0.013***        
> 0.022***        0.025***        0.017***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
FDI jobs c~d        0.010***        0.010***                        0.007***        0.006***                        
> 0.010***        0.007***                
                   (0.00)          (0.00)                          (0.00)          (0.00)                          (
> 0.00)          (0.00)                   
FDI jobs *~n       -0.002***       -0.002***                       -0.001*         -0.001*                         -
> 0.001**        -0.001**                 
                   (0.00)          (0.00)                          (0.00)          (0.00)                          (
> 0.00)          (0.00)                   
Age in years       -0.002***       -0.002***       -0.002***       -0.002***       -0.002***       -0.002***       -
> 0.002***       -0.002***       -0.002***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female             -0.001          -0.008          -0.008           0.001          -0.001          -0.001          -
> 0.008          -0.016**        -0.012*  
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Unemployed         -0.006          -0.007          -0.007          -0.001          -0.001          -0.001          -
> 0.015*         -0.004          -0.015** 
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Urban resi~t        0.056***        0.052***        0.050***        0.056***        0.052***        0.045***        
> 0.056***        0.062***        0.046***
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Political ~t                       -0.031***       -0.031***                       -0.023***       -0.023***        
>                -0.036***       -0.033***
                                   (0.00)          (0.00)                          (0.00)          (0.00)           
>                (0.00)          (0.00)   
Newspaper ~n                        0.008***        0.007***                        0.011***        0.011***        
>                 0.009***        0.016***
                                   (0.00)          (0.00)                          (0.00)          (0.00)           
>                (0.00)          (0.00)   
Black/Afri~n                        0.000                                           0.000                           
>                 0.000                   
                                      (.)                                             (.)                           
>                   (.)                   
White/Euro~n                        0.034                                           0.050                           
>                -0.054                   
                                   (0.06)                                          (0.05)                           
>                (0.05)                   
Colored/mi~e                       -0.044                                          -0.042                           
>                -0.046                   
                                   (0.04)                                          (0.03)                           
>                (0.03)                   
Arab/Leban~n                       -0.022                                           0.055                           
>                 0.085***                
                                   (0.05)                                          (0.04)                           
>                (0.01)                   
Other                               0.127                                          -0.179**                         
>                -0.031                   
                                   (0.09)                                          (0.09)                           
>                (0.09)                   
lnfdi_job~10                                        0.008***                                        0.007***        
>                                 0.006***
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
c.lnfdi_jo~l                                       -0.001***                                       -0.001           
>                                -0.001*  
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        18738           18490           18583           21063           20764           20859           
> 19705           19444           19536   
\# of coun~s           19              19              19              20              20              20           
>    19              19              19   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
Log likeli~d                                                                                                        
>                                         
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** Set 3: Type of Political Regime
. * Opinion: Country Is Not a Democracy
. xtreg demoqua c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,312
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0157                                         min =        207
     Between = 0.2057                                         avg =    1,015.6
     Overall = 0.0232                                         max =      2,146

                                                Wald chi2(9)      =     326.13
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0190974   .0046876     4.07   0.000       .00991    .0282849
               educlvl |   .0538105   .0063005     8.54   0.000     .0414618    .0661591
                       |
c.lnfdi_jobs#c.educlvl |  -.0041531   .0009551    -4.35   0.000     -.006025   -.0022813
                       |
                   age |  -.0008151   .0004773    -1.71   0.088    -.0017505    .0001203
                female |   -.055205   .0123772    -4.46   0.000    -.0794638   -.0309461
                unempl |  -.0333496   .0137887    -2.42   0.016     -.060375   -.0063242
                 urban |   .0766486   .0153879     4.98   0.000     .0464889    .1068084
                       |
                 round |
                    5  |    .158075   .0192469     8.21   0.000     .1203518    .1957981
                    6  |   .0705602   .0198274     3.56   0.000     .0316991    .1094212
                       |
                 _cons |   2.332207   .0595959    39.13   0.000     2.215401    2.449013
-----------------------+----------------------------------------------------------------
               sigma_u |   .2051467
               sigma_e |  .86293066
                   rho |  .05349348   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demoqua1

. xtreg demoqua c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,139
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0186                                         min =        205
     Between = 0.1980                                         avg =    1,007.0
     Overall = 0.0263                                         max =      2,130

                                                Wald chi2(11)     =     385.36
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0182777   .0047273     3.87   0.000     .0090123     .027543
               educlvl |   .0539733    .006476     8.33   0.000     .0412806    .0666661
                       |
c.lnfdi_jobs#c.educlvl |  -.0041256    .000961    -4.29   0.000    -.0060091   -.0022422
                       |
                   age |  -.0004586   .0004815    -0.95   0.341    -.0014022    .0004851
                female |  -.0645194    .012599    -5.12   0.000     -.089213   -.0398258
                unempl |   -.031379   .0138434    -2.27   0.023    -.0585116   -.0042464
                 urban |   .0684748   .0156761     4.37   0.000     .0377503    .0991993
              interest |   -.050375   .0061539    -8.19   0.000    -.0624364   -.0383136
                  news |   .0092748      .0047     1.97   0.048      .000063    .0184866
                       |
                 round |
                    5  |   .1568107   .0192935     8.13   0.000     .1189961    .1946253
                    6  |   .0677677   .0198699     3.41   0.001     .0288234    .1067121
                       |
                 _cons |   2.403199   .0574776    41.81   0.000     2.290545    2.515853
-----------------------+----------------------------------------------------------------
               sigma_u |  .18787816
               sigma_e |  .86112649
                   rho |  .04543836   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demoqua2

. xtreg demoqua c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,050
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0194                                         min =        123
     Between = 0.1524                                         avg =    1,002.5
     Overall = 0.0225                                         max =      2,130

                                                Wald chi2(15)     =     399.40
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0183512   .0047331     3.88   0.000     .0090745    .0276279
               educlvl |   .0538021   .0064771     8.31   0.000     .0411071     .066497
                       |
c.lnfdi_jobs#c.educlvl |  -.0041633   .0009622    -4.33   0.000    -.0060492   -.0022774
                       |
                   age |  -.0004778   .0004826    -0.99   0.322    -.0014237    .0004682
                female |  -.0641941   .0126199    -5.09   0.000    -.0889286   -.0394595
                unempl |  -.0322934   .0138713    -2.33   0.020    -.0594807   -.0051062
                 urban |   .0687728   .0157206     4.37   0.000      .037961    .0995846
              interest |  -.0503067   .0061625    -8.16   0.000     -.062385   -.0382285
                  news |   .0084655   .0047173     1.79   0.073    -.0007802    .0177112
                       |
                  race |
                    2  |   .1844248   .0974118     1.89   0.058    -.0064988    .3753485
                    3  |      .1387   .0613738     2.26   0.024     .0184096    .2589904
                    4  |  -.0319538   .0870061    -0.37   0.713    -.2024826     .138575
                    5  |   .2765099   .1631705     1.69   0.090    -.0432984    .5963182
                       |
                 round |
                    5  |   .1568751   .0193161     8.12   0.000     .1190163    .1947339
                    6  |   .0657902   .0199055     3.31   0.001     .0267762    .1048042
                       |
                 _cons |   2.404134    .068647    35.02   0.000     2.269588    2.538679
-----------------------+----------------------------------------------------------------
               sigma_u |  .23939863
               sigma_e |  .86102671
                   rho |  .07175815   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demoqua3

. xtreg demoqua c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,139
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0186                                         min =        205
     Between = 0.1851                                         avg =    1,007.0
     Overall = 0.0261                                         max =      2,130

                                                Wald chi2(11)     =     385.90
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0197842   .0052311     3.78   0.000     .0095314    .0300369
                educlvl |   .0528832   .0061924     8.54   0.000     .0407464      .06502
                        |
c.lnfdi_capex#c.educlvl |  -.0045921   .0010547    -4.35   0.000    -.0066594   -.0025249
                        |
                    age |  -.0004395   .0004815    -0.91   0.361    -.0013832    .0005042
                 female |  -.0645749   .0125993    -5.13   0.000     -.089269   -.0398808
                 unempl |  -.0311916   .0138439    -2.25   0.024    -.0583251    -.004058
                  urban |   .0683018    .015756     4.33   0.000     .0374205    .0991831
               interest |  -.0503686   .0061543    -8.18   0.000    -.0624307   -.0383064
                   news |   .0092704   .0047005     1.97   0.049     .0000577    .0184832
                        |
                  round |
                     5  |   .1578275   .0192317     8.21   0.000     .1201341     .195521
                     6  |   .0695293   .0198776     3.50   0.000       .03057    .1084886
                        |
                  _cons |   2.409273   .0563304    42.77   0.000     2.298867    2.519678
------------------------+----------------------------------------------------------------
                sigma_u |  .18433603
                sigma_e |  .86111229
                    rho |  .04381693   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store demoqua4

. xtoprobit demoqua c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -26160.975  
Iteration 1:   log likelihood = -25831.947  
Iteration 2:   log likelihood = -25831.912  
Iteration 3:   log likelihood = -25831.912  

Refining starting values:

Grid node 0:   log likelihood = -25029.431

Fitting full model:

Iteration 0:   log likelihood = -25029.431  (not concave)
Iteration 1:   log likelihood = -25020.198  (not concave)
Iteration 2:   log likelihood = -25015.996  (not concave)
Iteration 3:   log likelihood = -25012.434  (not concave)
Iteration 4:   log likelihood = -25009.552  
Iteration 5:   log likelihood = -24949.468  
Iteration 6:   log likelihood = -24928.142  
Iteration 7:   log likelihood = -24927.996  
Iteration 8:   log likelihood = -24927.996  

Random-effects ordered probit regression            Number of obs    =  20,139
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     205
                                                                 avg = 1,007.0
                                                                 max =   2,130

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  381.62
Log likelihood = -24927.996                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .023138   .0059087     3.92   0.000     .0115572    .0347187
               educlvl |   .0674269   .0080936     8.33   0.000     .0515637    .0832901
                       |
c.lnfdi_jobs#c.educlvl |  -.0052394   .0012004    -4.36   0.000    -.0075922   -.0028867
                       |
                   age |  -.0005951   .0006013    -0.99   0.322    -.0017737    .0005835
                female |  -.0816032   .0157243    -5.19   0.000    -.1124222   -.0507842
                unempl |  -.0396457   .0172783    -2.29   0.022    -.0735105   -.0057809
                 urban |   .0846611   .0195801     4.32   0.000     .0462848    .1230375
              interest |  -.0633085   .0076967    -8.23   0.000    -.0783938   -.0482231
                  news |   .0122307   .0058717     2.08   0.037     .0007224     .023739
                       |
                 round |
                    5  |   .1966151   .0241068     8.16   0.000     .1493667    .2438636
                    6  |   .0864559   .0248205     3.48   0.000     .0378086    .1351031
-----------------------+----------------------------------------------------------------
                 /cut1 |   -.999314    .090596                     -1.176879   -.8217491
                 /cut2 |   .1182582   .0903845                      -.058892    .2954085
                 /cut3 |   1.241348   .0906286                       1.06372    1.418977
-----------------------+----------------------------------------------------------------
             /sigma2_u |   .1147278   .0368343                      .0611482    .2152552
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 1807.83      Prob >= chibar2 = 0.0000

. est store demoqua5

. xtreg demoqua c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,139
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0184                                         min =        205
     Between = 0.1984                                         avg =    1,007.0
     Overall = 0.0260                                         max =      2,130

                                                Wald chi2(11)     =     378.96
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                  demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0148841   .0040677     3.66   0.000     .0069115    .0228567
                  educlvl |   .0407865   .0047232     8.64   0.000     .0315291    .0500439
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0026616   .0008089    -3.29   0.001    -.0042471   -.0010761
                          |
                      age |  -.0004519   .0004816    -0.94   0.348    -.0013957     .000492
                   female |  -.0652259   .0125964    -5.18   0.000    -.0899144   -.0405374
                   unempl |  -.0319139   .0138423    -2.31   0.021    -.0590442   -.0047836
                    urban |   .0620659   .0161101     3.85   0.000     .0304907    .0936412
                 interest |  -.0504174   .0061518    -8.20   0.000    -.0624747   -.0383602
                     news |   .0093989   .0046968     2.00   0.045     .0001933    .0186045
                          |
                    round |
                       5  |   .1562985   .0190301     8.21   0.000     .1190003    .1935967
                       6  |   .0656479   .0193901     3.39   0.001      .027644    .1036519
                          |
                    _cons |   2.455657   .0613904    40.00   0.000     2.335334     2.57598
--------------------------+----------------------------------------------------------------
                  sigma_u |  .22598999
                  sigma_e |   .8612401
                      rho |  .06441863   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store demoqua6

. * Dissatisfaction with How Democracy Works
. xtreg demosat c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,667
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0135                                         min =        210
     Between = 0.0042                                         avg =    1,033.3
     Overall = 0.0144                                         max =      2,180

                                                Wald chi2(9)      =     282.62
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0172418   .0050495     3.41   0.001     .0073449    .0271387
               educlvl |   .0467388   .0068321     6.84   0.000      .033348    .0601295
                       |
c.lnfdi_jobs#c.educlvl |   -.003322    .001034    -3.21   0.001    -.0053486   -.0012953
                       |
                   age |  -.0014946   .0005192    -2.88   0.004    -.0025123   -.0004769
                female |  -.0551497   .0134907    -4.09   0.000    -.0815911   -.0287083
                unempl |  -.0111592   .0150152    -0.74   0.457    -.0405885      .01827
                 urban |    .114158   .0167572     6.81   0.000     .0813145    .1470015
                       |
                 round |
                    5  |   .0724506   .0210037     3.45   0.001     .0312841    .1136171
                    6  |  -.0472577   .0215928    -2.19   0.029    -.0895788   -.0049366
                       |
                 _cons |    1.62818   .0783439    20.78   0.000     1.474628    1.781731
-----------------------+----------------------------------------------------------------
               sigma_u |  .29811204
               sigma_e |  .94908747
                   rho |  .08980133   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demosat1

. xtreg demosat c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,486
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0216                                         min =        208
     Between = 0.0328                                         avg =    1,024.3
     Overall = 0.0238                                         max =      2,161

                                                Wald chi2(11)     =     453.15
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .015368   .0050823     3.02   0.002     .0054068    .0253292
               educlvl |   .0461267   .0070132     6.58   0.000      .032381    .0598724
                       |
c.lnfdi_jobs#c.educlvl |  -.0032212   .0010383    -3.10   0.002    -.0052563   -.0011862
                       |
                   age |  -.0008282   .0005226    -1.58   0.113    -.0018525    .0001961
                female |  -.0722404   .0137055    -5.27   0.000    -.0991027   -.0453781
                unempl |  -.0113564   .0150483    -0.75   0.450    -.0408504    .0181377
                 urban |   .1014529   .0170336     5.96   0.000     .0680677    .1348381
              interest |  -.0869897   .0066793   -13.02   0.000     -.100081   -.0738984
                  news |   .0206825    .005128     4.03   0.000     .0106318    .0307331
                       |
                 round |
                    5  |   .0658503   .0210179     3.13   0.002      .024656    .1070446
                    6  |  -.0550626   .0216036    -2.55   0.011    -.0974049   -.0127203
                       |
                 _cons |   1.748807   .0733794    23.83   0.000     1.604986    1.892628
-----------------------+----------------------------------------------------------------
               sigma_u |  .26800671
               sigma_e |  .94534811
                   rho |  .07439338   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demosat2

. xtreg demosat c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,397
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0223                                         min =        126
     Between = 0.0658                                         avg =    1,019.9
     Overall = 0.0269                                         max =      2,161

                                                Wald chi2(15)     =     466.38
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0155822   .0050828     3.07   0.002     .0056202    .0255443
               educlvl |    .045864   .0070076     6.54   0.000     .0321294    .0595986
                       |
c.lnfdi_jobs#c.educlvl |  -.0032375   .0010386    -3.12   0.002    -.0052731   -.0012018
                       |
                   age |   -.000851   .0005234    -1.63   0.104    -.0018768    .0001748
                female |  -.0717186   .0137154    -5.23   0.000    -.0986002   -.0448369
                unempl |  -.0096356   .0150635    -0.64   0.522    -.0391595    .0198883
                 urban |   .1034947   .0170597     6.07   0.000     .0700584    .1369311
              interest |  -.0867457   .0066817   -12.98   0.000    -.0998417   -.0736498
                  news |   .0206889   .0051405     4.02   0.000     .0106136    .0307641
                       |
                  race |
                    2  |   .1512804   .1064591     1.42   0.155    -.0573757    .3599365
                    3  |   .0164053   .0672958     0.24   0.807    -.1154921    .1483027
                    4  |  -.0703195    .101177    -0.70   0.487    -.2686229    .1279839
                    5  |    .316642   .1758096     1.80   0.072    -.0279386    .6612225
                       |
                 round |
                    5  |   .0682026   .0210189     3.24   0.001     .0270063    .1093988
                    6  |  -.0570661   .0216164    -2.64   0.008    -.0994334   -.0146989
                       |
                 _cons |   1.751927   .0852087    20.56   0.000     1.584921    1.918933
-----------------------+----------------------------------------------------------------
               sigma_u |  .31779787
               sigma_e |  .94422852
                   rho |  .10175222   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demosat3

. xtreg demosat c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,486
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0216                                         min =        208
     Between = 0.0387                                         avg =    1,024.3
     Overall = 0.0240                                         max =      2,161

                                                Wald chi2(11)     =     451.81
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0148555   .0056244     2.64   0.008     .0038319     .025879
                educlvl |   .0443149    .006709     6.61   0.000     .0311655    .0574643
                        |
c.lnfdi_capex#c.educlvl |  -.0033938   .0011399    -2.98   0.003    -.0056279   -.0011598
                        |
                    age |  -.0008074   .0005226    -1.54   0.122    -.0018316    .0002169
                 female |   -.072178   .0137044    -5.27   0.000    -.0990381   -.0453179
                 unempl |   -.011391   .0150477    -0.76   0.449     -.040884    .0181019
                  urban |   .1026904   .0171208     6.00   0.000     .0691342    .1362466
               interest |  -.0870034   .0066797   -13.03   0.000    -.1000954   -.0739114
                   news |    .020896   .0051289     4.07   0.000     .0108436    .0309484
                        |
                  round |
                     5  |   .0680207   .0209512     3.25   0.001      .026957    .1090844
                     6  |    -.05204   .0216141    -2.41   0.016    -.0944028   -.0096771
                        |
                  _cons |    1.75977    .077511    22.70   0.000     1.607851    1.911689
------------------------+----------------------------------------------------------------
                sigma_u |  .29283815
                sigma_e |  .94537983
                    rho |  .08754919   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store demosat4

. xtoprobit demosat c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood =  -28841.08  
Iteration 1:   log likelihood = -28529.981  
Iteration 2:   log likelihood = -28529.944  
Iteration 3:   log likelihood = -28529.944  

Refining starting values:

Grid node 0:   log likelihood =  -27598.41

Fitting full model:

Iteration 0:   log likelihood =  -27598.41  (not concave)
Iteration 1:   log likelihood = -27593.814  (not concave)
Iteration 2:   log likelihood = -27589.216  (not concave)
Iteration 3:   log likelihood = -27585.331  (not concave)
Iteration 4:   log likelihood = -27582.082  (not concave)
Iteration 5:   log likelihood = -27579.393  (not concave)
Iteration 6:   log likelihood = -27577.219  
Iteration 7:   log likelihood = -27532.517  
Iteration 8:   log likelihood = -27474.053  
Iteration 9:   log likelihood = -27473.985  
Iteration 10:  log likelihood = -27473.985  

Random-effects ordered probit regression            Number of obs    =  20,486
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     208
                                                                 avg = 1,024.3
                                                                 max =   2,161

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  442.99
Log likelihood = -27473.985                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
               demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0174404   .0056888     3.07   0.002     .0062906    .0285903
               educlvl |    .051695   .0078472     6.59   0.000     .0363146    .0670753
                       |
c.lnfdi_jobs#c.educlvl |  -.0036305    .001161    -3.13   0.002     -.005906   -.0013549
                       |
                   age |  -.0009712   .0005853    -1.66   0.097    -.0021183    .0001759
                female |   -.081706   .0153293    -5.33   0.000    -.1117509    -.051661
                unempl |  -.0118755   .0168229    -0.71   0.480    -.0448479    .0210968
                 urban |   .1116489   .0190666     5.86   0.000      .074279    .1490188
              interest |  -.0969082    .007498   -12.92   0.000     -.111604   -.0822125
                  news |      .0239    .005738     4.17   0.000     .0126538    .0351462
                       |
                 round |
                    5  |   .0736518   .0235412     3.13   0.002     .0275118    .1197918
                    6  |  -.0556039    .024203    -2.30   0.022     -.103041   -.0081668
-----------------------+----------------------------------------------------------------
                 /cut1 |  -1.334088   .0929566                      -1.51628   -1.151897
                 /cut2 |  -.1983547   .0925562                     -.3797615   -.0169478
                 /cut3 |   .7220226   .0926215                      .5404878    .9035573
                 /cut4 |   2.054984   .0939541                      1.870837    2.239131
-----------------------+----------------------------------------------------------------
             /sigma2_u |    .125837   .0402034                      .0672759    .2353732
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 2111.92      Prob >= chibar2 = 0.0000

. est store demosat5

. xtreg demosat c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,486
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0215                                         min =        208
     Between = 0.0274                                         avg =    1,024.3
     Overall = 0.0234                                         max =      2,161

                                                Wald chi2(11)     =     450.13
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-------------------------------------------------------------------------------------------
                  demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0117697   .0043819     2.69   0.007     .0031812    .0203581
                  educlvl |   .0356906   .0051187     6.97   0.000     .0256581     .045723
                          |
c.lnfdi_jobs_10#c.educlvl |  -.0020227   .0008751    -2.31   0.021    -.0037379   -.0003075
                          |
                      age |  -.0008122   .0005228    -1.55   0.120    -.0018368    .0002125
                   female |  -.0728586   .0137053    -5.32   0.000    -.0997205   -.0459968
                   unempl |   -.011489   .0150494    -0.76   0.445    -.0409853    .0180073
                    urban |   .0971398   .0175118     5.55   0.000     .0628174    .1314623
                 interest |  -.0871465   .0066776   -13.05   0.000    -.1002344   -.0740586
                     news |   .0208049   .0051227     4.06   0.000     .0107645    .0308452
                          |
                    round |
                       5  |   .0675229   .0207259     3.26   0.001      .026901    .1081449
                       6  |  -.0540001   .0210753    -2.56   0.010    -.0953068   -.0126933
                          |
                    _cons |   1.791408   .0721991    24.81   0.000     1.649901    1.932916
--------------------------+----------------------------------------------------------------
                  sigma_u |  .27531134
                  sigma_e |  .94541431
                      rho |  .07817242   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store demosat6

. * Preferences for Democracy
. xtreg demosup c.lnfdi_jobs##c.educlvl $C1 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     17,777
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0021                                         min =        192
     Between = 0.0935                                         avg =      888.9
     Overall = 0.0010                                         max =      2,006

                                                Wald chi2(9)      =      36.29
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0070775   .0021016     3.37   0.001     .0029584    .0111966
               educlvl |   .0122595   .0027919     4.39   0.000     .0067875    .0177314
                       |
c.lnfdi_jobs#c.educlvl |  -.0013038   .0004234    -3.08   0.002    -.0021336   -.0004739
                       |
                   age |    .000703   .0002126     3.31   0.001     .0002863    .0011198
                female |   .0010274   .0055145     0.19   0.852     -.009781    .0118357
                unempl |   .0018661   .0061511     0.30   0.762    -.0101897     .013922
                 urban |   .0061966   .0068292     0.91   0.364    -.0071883    .0195815
                       |
                 round |
                    5  |  -.0114964   .0085895    -1.34   0.181    -.0283314    .0053387
                    6  |  -.0150742   .0088459    -1.70   0.088    -.0324118    .0022634
                       |
                 _cons |   .7388327   .0271265    27.24   0.000     .6856657    .7919998
-----------------------+----------------------------------------------------------------
               sigma_u |  .09475923
               sigma_e |  .35979672
                   rho |  .06486396   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demosup1

. xtreg demosup c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     17,634
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0022                                         min =        190
     Between = 0.1183                                         avg =      881.7
     Overall = 0.0010                                         max =      1,990

                                                Wald chi2(11)     =      38.52
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0001

----------------------------------------------------------------------------------------
               demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0071492   .0021238     3.37   0.001     .0029865    .0113118
               educlvl |   .0118591   .0028754     4.12   0.000     .0062234    .0174948
                       |
c.lnfdi_jobs#c.educlvl |  -.0013251   .0004271    -3.10   0.002    -.0021621   -.0004881
                       |
                   age |   .0007002    .000215     3.26   0.001     .0002789    .0011215
                female |   .0006325   .0056377     0.11   0.911    -.0104171    .0116821
                unempl |   .0028936   .0061919     0.47   0.640    -.0092422    .0150294
                 urban |   .0068123   .0069703     0.98   0.328    -.0068492    .0204738
              interest |   .0006507   .0027789     0.23   0.815    -.0047959    .0060973
                  news |   .0018242   .0020981     0.87   0.385     -.002288    .0059364
                       |
                 round |
                    5  |  -.0112223   .0086324    -1.30   0.194    -.0281414    .0056968
                    6  |  -.0148911   .0088882    -1.68   0.094    -.0323117    .0025295
                       |
                 _cons |   .7361241    .023846    30.87   0.000     .6893869    .7828614
-----------------------+----------------------------------------------------------------
               sigma_u |  .07216385
               sigma_e |  .36011564
                   rho |  .03860617   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demosup2

. xtreg demosup c.lnfdi_jobs##c.educlvl $C1 $C2 $C3 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     17,549
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0028                                         min =        110
     Between = 0.3876                                         avg =      877.5
     Overall = 0.0103                                         max =      1,990

                                                Wald chi2(15)     =      61.30
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0072505   .0021238     3.41   0.001     .0030879    .0114131
               educlvl |   .0119333   .0028724     4.15   0.000     .0063034    .0175631
                       |
c.lnfdi_jobs#c.educlvl |  -.0013331   .0004272    -3.12   0.002    -.0021705   -.0004958
                       |
                   age |   .0007567   .0002153     3.51   0.000     .0003348    .0011787
                female |   .0001511   .0056434     0.03   0.979    -.0109098     .011212
                unempl |   .0023372   .0061996     0.38   0.706    -.0098138    .0144882
                 urban |   .0075322   .0069662     1.08   0.280    -.0061213    .0211857
              interest |   .0002338   .0027789     0.08   0.933    -.0052127    .0056803
                  news |   .0022529   .0020998     1.07   0.283    -.0018627    .0063684
                       |
                  race |
                    2  |  -.0602453   .0428145    -1.41   0.159    -.1441601    .0236695
                    3  |  -.1223164   .0269009    -4.55   0.000    -.1750413   -.0695916
                    4  |   -.037569   .0253728    -1.48   0.139    -.0872987    .0121607
                    5  |    .032848   .0694656     0.47   0.636     -.103302    .1689981
                       |
                 round |
                    5  |  -.0124781   .0086126    -1.45   0.147    -.0293584    .0044023
                    6  |  -.0176067   .0088743    -1.98   0.047    -.0349999   -.0002135
                       |
                 _cons |   .7461681   .0212427    35.13   0.000     .7045332     .787803
-----------------------+----------------------------------------------------------------
               sigma_u |  .04937928
               sigma_e |  .35931373
                   rho |  .01853601   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store demosup3

. xtreg demosup c.lnfdi_capex##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     17,634
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0019                                         min =        190
     Between = 0.1513                                         avg =      881.7
     Overall = 0.0006                                         max =      1,990

                                                Wald chi2(11)     =      32.16
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0007

-----------------------------------------------------------------------------------------
                demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0053672   .0023529     2.28   0.023     .0007557    .0099788
                educlvl |   .0087726   .0027489     3.19   0.001     .0033847    .0141604
                        |
c.lnfdi_capex#c.educlvl |  -.0009209   .0004683    -1.97   0.049    -.0018388   -3.11e-06
                        |
                    age |   .0007051    .000215     3.28   0.001     .0002837    .0011265
                 female |    .000506   .0056389     0.09   0.928     -.010546     .011558
                 unempl |   .0027299   .0061932     0.44   0.659    -.0094086    .0148683
                  urban |   .0072059   .0070077     1.03   0.304     -.006529    .0209407
               interest |   .0005762   .0027798     0.21   0.836    -.0048721    .0060245
                   news |   .0019015   .0020984     0.91   0.365    -.0022114    .0060143
                        |
                  round |
                     5  |  -.0103494   .0086047    -1.20   0.229    -.0272144    .0065156
                     6  |   -.014166   .0088922    -1.59   0.111    -.0315944    .0032623
                        |
                  _cons |   .7492924   .0232847    32.18   0.000     .7036552    .7949296
------------------------+----------------------------------------------------------------
                sigma_u |  .07025523
                sigma_e |  .36017947
                    rho |   .0366524   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store demosup4

. xtprobit demosup c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood =  -7833.167  
Iteration 1:   log likelihood = -7802.5805  
Iteration 2:   log likelihood = -7802.5317  
Iteration 3:   log likelihood = -7802.5317  

Fitting full model:

rho =  0.0     log likelihood = -7802.5317
rho =  0.1     log likelihood = -7490.4042
rho =  0.2     log likelihood =  -7525.883

Iteration 0:   log likelihood = -7480.7796  
Iteration 1:   log likelihood = -7462.5148  
Iteration 2:   log likelihood = -7462.5099  
Iteration 3:   log likelihood = -7462.5099  

Random-effects probit regression                     Number of obs    = 17,634
Group variable: cname                                Number of groups =     20

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    190
                                                                  avg =  881.7
                                                                  max =  1,990

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(11)    =  36.56
Log likelihood = -7462.5099                          Prob > chi2      = 0.0001

----------------------------------------------------------------------------------------
               demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0288271   .0088155     3.27   0.001     .0115491    .0461051
               educlvl |    .047717   .0121431     3.93   0.000      .023917     .071517
                       |
c.lnfdi_jobs#c.educlvl |  -.0054236   .0017839    -3.04   0.002      -.00892   -.0019273
                       |
                   age |   .0030783   .0009252     3.33   0.001     .0012649    .0048917
                female |   .0021425   .0240992     0.09   0.929     -.045091    .0493761
                unempl |   .0160907   .0264496     0.61   0.543    -.0357496     .067931
                 urban |    .027594   .0294955     0.94   0.350     -.030216    .0854041
              interest |   .0018654   .0119715     0.16   0.876    -.0215984    .0253292
                  news |   .0091661   .0089264     1.03   0.304    -.0083293    .0266615
                       |
                 round |
                    5  |  -.0486699   .0381598    -1.28   0.202    -.1234617    .0261218
                    6  |   -.061755    .039237    -1.57   0.116    -.1386581    .0151482
                       |
                 _cons |   .6224175   .1076474     5.78   0.000     .4114326    .8334025
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.111447   .3280024                     -2.754319   -1.468574
-----------------------+----------------------------------------------------------------
               sigma_u |   .3479407   .0570627                      .2522941    .4798476
                   rho |   .1079892   .0315957                      .0598432    .1871595
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 680.04                 Prob >= chibar2 = 0.000

. est store demosup5

. xtreg demosup c.lnfdi_jobs_10##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     17,634
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0019                                         min =        190
     Between = 0.1905                                         avg =      881.7
     Overall = 0.0006                                         max =      1,990

                                                Wald chi2(11)     =      32.71
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0006

-------------------------------------------------------------------------------------------
                  demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
--------------------------+----------------------------------------------------------------
            lnfdi_jobs_10 |   .0043194   .0018287     2.36   0.018     .0007352    .0079035
                  educlvl |   .0074991   .0020844     3.60   0.000     .0034138    .0115844
                          |
c.lnfdi_jobs_10#c.educlvl |   -.000806   .0003587    -2.25   0.025    -.0015092   -.0001029
                          |
                      age |   .0007152   .0002151     3.33   0.001     .0002937    .0011367
                   female |   .0004263   .0056382     0.08   0.940    -.0106243    .0114769
                   unempl |   .0027805   .0061931     0.45   0.653    -.0093579    .0149188
                    urban |   .0073171   .0071584     1.02   0.307    -.0067131    .0213473
                 interest |   .0004568    .002778     0.16   0.869     -.004988    .0059016
                     news |    .002013   .0020962     0.96   0.337    -.0020954    .0061214
                          |
                    round |
                       5  |  -.0089947   .0085105    -1.06   0.291     -.025675    .0076857
                       6  |  -.0123539   .0086748    -1.42   0.154    -.0293562    .0046483
                          |
                    _cons |   .7564357   .0222556    33.99   0.000     .7128155     .800056
--------------------------+----------------------------------------------------------------
                  sigma_u |  .07111272
                  sigma_e |  .36017415
                      rho |  .03751987   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------

. est store demosup6

. 
. *** TABLE A-8: Type of Political Regime - Main Models
. esttab demoqua2 demoqua4 demoqua5 demosat2 demosat4 demosat5 demosup2 demosup4 demosup5, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl lnfdi_capex c.lnfdi_capex#c.educlvl age female unempl urba
> n interest news) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 lnfdi_capex "FDI capital expenditure" ///
>                 c.lnfdi_jobs#c.educlvl "FDI jobs * education" ///
>                 c.lnfdi_capex#c.educlvl "FDI capex * education" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                  demoqua         demoqua         demoqua         demosat         demosat         demosat         de
> mosup         demosup         demosup   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
Education ~l        0.054***        0.053***        0.067***        0.046***        0.044***        0.052***        
> 0.012***        0.009***        0.048***
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.00)          (0.00)          (0.01)   
FDI jobs c~d        0.018***                        0.023***        0.015***                        0.017***        
> 0.007***                        0.029***
                   (0.00)                          (0.01)          (0.01)                          (0.01)          (
> 0.00)                          (0.01)   
FDI jobs *~n       -0.004***                       -0.005***       -0.003***                       -0.004***       -
> 0.001***                       -0.005***
                   (0.00)                          (0.00)          (0.00)                          (0.00)          (
> 0.00)                          (0.00)   
FDI capita~e                        0.020***                                        0.015***                        
>                 0.005**                 
                                   (0.01)                                          (0.01)                           
>                (0.00)                   
FDI capex ~n                       -0.005***                                       -0.003***                        
>                -0.001**                 
                                   (0.00)                                          (0.00)                           
>                (0.00)                   
Age in years       -0.000          -0.000          -0.001          -0.001          -0.001          -0.001*          
> 0.001***        0.001***        0.003***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female             -0.065***       -0.065***       -0.082***       -0.072***       -0.072***       -0.082***        
> 0.001           0.001           0.002   
                   (0.01)          (0.01)          (0.02)          (0.01)          (0.01)          (0.02)          (
> 0.01)          (0.01)          (0.02)   
Unemployed         -0.031**        -0.031**        -0.040**        -0.011          -0.011          -0.012           
> 0.003           0.003           0.016   
                   (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (0.02)          (
> 0.01)          (0.01)          (0.03)   
Urban resi~t        0.068***        0.068***        0.085***        0.101***        0.103***        0.112***        
> 0.007           0.007           0.028   
                   (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (
> 0.01)          (0.01)          (0.03)   
Political ~t       -0.050***       -0.050***       -0.063***       -0.087***       -0.087***       -0.097***        
> 0.001           0.001           0.002   
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.00)          (0.00)          (0.01)   
Newspaper ~n        0.009**         0.009**         0.012**         0.021***        0.021***        0.024***        
> 0.002           0.002           0.009   
                   (0.00)          (0.00)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.00)          (0.00)          (0.01)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
lnsig2u                                                                                                             
>                                -2.111***
                                                                                                                    
>                                (0.33)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        20139           20139           20139           20486           20486           20486           
> 17634           17634           17634   
\# of coun~s           20              20              20              20              20              20           
>    20              20              20   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.001           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** TABLE A-9: Type of Political Regime - Robustness Tests
. esttab demoqua1 demoqua3 demoqua6 demosat1 demosat3 demosat6 demosup1 demosup3 demosup6, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p ll, fmt(0 0 3 2) labels("\# respondents" "\# of countries" "Prob > Chi2" "Log likelihood")) 
> ///
>         order(educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl age female unempl urban interest news *race) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 lnfdi_capex "FDI capital expenditure" ///
>                 c.lnfdi_jobs#c.educlvl "FDI jobs * education" ///
>                 c.lnfdi_capex#c.educlvl "FDI capex * education" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption" 1.race "Black/African" ///
>                 2.race "White/European" 3.race "Colored/mixed race" ///
>                 4.race "Arab/Lebanese/North African" 5.race "Other")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                  demoqua         demoqua         demoqua         demosat         demosat         demosat         de
> mosup         demosup         demosup   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Education ~l        0.054***        0.054***        0.041***        0.047***        0.046***        0.036***        
> 0.012***        0.012***        0.007***
                   (0.01)          (0.01)          (0.00)          (0.01)          (0.01)          (0.01)          (
> 0.00)          (0.00)          (0.00)   
FDI jobs c~d        0.019***        0.018***                        0.017***        0.016***                        
> 0.007***        0.007***                
                   (0.00)          (0.00)                          (0.01)          (0.01)                          (
> 0.00)          (0.00)                   
FDI jobs *~n       -0.004***       -0.004***                       -0.003***       -0.003***                       -
> 0.001***       -0.001***                
                   (0.00)          (0.00)                          (0.00)          (0.00)                          (
> 0.00)          (0.00)                   
Age in years       -0.001*         -0.000          -0.000          -0.001***       -0.001          -0.001           
> 0.001***        0.001***        0.001***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female             -0.055***       -0.064***       -0.065***       -0.055***       -0.072***       -0.073***        
> 0.001           0.000           0.000   
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Unemployed         -0.033**        -0.032**        -0.032**        -0.011          -0.010          -0.011           
> 0.002           0.002           0.003   
                   (0.01)          (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (
> 0.01)          (0.01)          (0.01)   
Urban resi~t        0.077***        0.069***        0.062***        0.114***        0.103***        0.097***        
> 0.006           0.008           0.007   
                   (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (
> 0.01)          (0.01)          (0.01)   
Political ~t                       -0.050***       -0.050***                       -0.087***       -0.087***        
>                 0.000           0.000   
                                   (0.01)          (0.01)                          (0.01)          (0.01)           
>                (0.00)          (0.00)   
Newspaper ~n                        0.008*          0.009**                         0.021***        0.021***        
>                 0.002           0.002   
                                   (0.00)          (0.00)                          (0.01)          (0.01)           
>                (0.00)          (0.00)   
Black/Afri~n                        0.000                                           0.000                           
>                 0.000                   
                                      (.)                                             (.)                           
>                   (.)                   
White/Euro~n                        0.184*                                          0.151                           
>                -0.060                   
                                   (0.10)                                          (0.11)                           
>                (0.04)                   
Colored/mi~e                        0.139**                                         0.016                           
>                -0.122***                
                                   (0.06)                                          (0.07)                           
>                (0.03)                   
Arab/Leban~n                       -0.032                                          -0.070                           
>                -0.038                   
                                   (0.09)                                          (0.10)                           
>                (0.03)                   
Other                               0.277*                                          0.317*                          
>                 0.033                   
                                   (0.16)                                          (0.18)                           
>                (0.07)                   
lnfdi_job~10                                        0.015***                                        0.012***        
>                                 0.004** 
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
c.lnfdi_jo~l                                       -0.003***                                       -0.002**         
>                                -0.001** 
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        20312           20050           20139           20667           20397           20486           
> 17777           17549           17634   
\# of coun~s           20              20              20              20              20              20           
>    20              20              20   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.001   
Log likeli~d                                                                                                        
>                                         
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** TABLE 1: FDI and Political Preferences, Afrobarometer
. esttab problem2 condlive2 condecon2 trustelec2 trustlead2 trustpart2 demoqua2 demosat2 demosup2, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl age female unempl urban interest news) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 c.lnfdi_jobs#c.educlvl "FDI jobs * education" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                  problem        condlive        condecon       trustelec       trustlead       trustpart         de
> moqua         demosat         demosup   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Education ~l        0.024***       -0.043***        0.014*          0.019***        0.015***        0.020***        
> 0.054***        0.046***        0.012***
                   (0.00)          (0.01)          (0.01)          (0.00)          (0.00)          (0.00)          (
> 0.01)          (0.01)          (0.00)   
FDI jobs c~d        0.013***        0.016***        0.026***        0.010***        0.006***        0.009***        
> 0.018***        0.015***        0.007***
                   (0.00)          (0.01)          (0.01)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.01)          (0.00)   
FDI jobs *~n       -0.003***       -0.002*         -0.002**        -0.002***       -0.001*         -0.001**        -
> 0.004***       -0.003***       -0.001***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Age in years       -0.002***        0.005***        0.003***       -0.002***       -0.002***       -0.002***       -
> 0.000          -0.001           0.001***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female              0.003          -0.069***       -0.035**        -0.008          -0.001          -0.012*         -
> 0.065***       -0.072***        0.001   
                   (0.01)          (0.02)          (0.02)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Unemployed         -0.029***       -0.150***       -0.071***       -0.007          -0.001          -0.015*         -
> 0.031**        -0.011           0.003   
                   (0.01)          (0.02)          (0.02)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.02)          (0.01)   
Urban resi~t        0.037***       -0.048**         0.055***        0.052***        0.051***        0.045***        
> 0.068***        0.101***        0.007   
                   (0.01)          (0.02)          (0.02)          (0.01)          (0.01)          (0.01)          (
> 0.02)          (0.02)          (0.01)   
Political ~t       -0.008**        -0.026***       -0.046***       -0.031***       -0.023***       -0.033***       -
> 0.050***       -0.087***        0.001   
                   (0.00)          (0.01)          (0.01)          (0.00)          (0.00)          (0.00)          (
> 0.01)          (0.01)          (0.00)   
Newspaper ~n        0.005**        -0.066***       -0.016***        0.007***        0.011***        0.016***        
> 0.009**         0.021***        0.002   
                   (0.00)          (0.01)          (0.01)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.01)          (0.00)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        21644           21859           21667           18583           20859           19536           
> 20139           20486           17634   
\# of coun~s           20              20              20              19              20              19           
>    20              20              20   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** FIGURE 1: Self-Reported Economic Problems
. xtreg problem c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,644
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0110                                         min =        218
     Between = 0.0209                                         avg =    1,082.2
     Overall = 0.0074                                         max =      2,334

                                                Wald chi2(11)     =     238.82
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0134041   .0024538     5.46   0.000     .0085948    .0182135
               educlvl |   .0238654   .0034469     6.92   0.000     .0171096    .0306212
                       |
c.lnfdi_jobs#c.educlvl |  -.0030808   .0005086    -6.06   0.000    -.0040776   -.0020841
                       |
                   age |  -.0015513   .0002578    -6.02   0.000    -.0020565   -.0010461
                female |   .0029886   .0068484     0.44   0.663    -.0104339    .0164111
                unempl |  -.0288069   .0074873    -3.85   0.000    -.0434816   -.0141321
                 urban |   .0368757   .0084071     4.39   0.000     .0203981    .0533532
              interest |  -.0075861   .0033075    -2.29   0.022    -.0140687   -.0011035
                  news |   .0051705   .0025735     2.01   0.045     .0001265    .0102145
                       |
                 round |
                    5  |  -.0509178   .0104697    -4.86   0.000     -.071438   -.0303976
                    6  |  -.0613164   .0106789    -5.74   0.000    -.0822467   -.0403862
                       |
                 _cons |   .4914222   .0335938    14.63   0.000     .4255795    .5572649
-----------------------+----------------------------------------------------------------
               sigma_u |  .11847009
               sigma_e |  .48420753
                   rho |  .05648132   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish 

Average marginal effects                                Number of obs = 21,644
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0134041   .0024538     5.46   0.000     .0085948    .0182135
          2  |   .0103233   .0020444     5.05   0.000     .0063164    .0143302
          3  |   .0072425   .0016897     4.29   0.000     .0039307    .0105542
          4  |   .0041616    .001431     2.91   0.004     .0013568    .0069664
          5  |   .0010808   .0013259     0.82   0.415    -.0015179    .0036795
          6  |      -.002    .001409    -1.42   0.156    -.0047617    .0007616
          7  |  -.0050809   .0016523    -3.08   0.002    -.0083192   -.0018425
          8  |  -.0081617    .001998    -4.08   0.000    -.0120776   -.0042457
          9  |  -.0112425   .0024023    -4.68   0.000    -.0159509   -.0065341
         10  |  -.0143234   .0028403    -5.04   0.000    -.0198903   -.0087564
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Self-Reported Economic Problems", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_problem, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. xtreg condlive c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0442                                         min =        219
     Between = 0.2551                                         avg =    1,093.0
     Overall = 0.0532                                         max =      2,337

                                                Wald chi2(11)     =    1013.24
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0155547   .0055124     2.82   0.005     .0047505    .0263589
               educlvl |  -.0434118   .0077559    -5.60   0.000     -.058613   -.0282105
                       |
c.lnfdi_jobs#c.educlvl |  -.0019371   .0011436    -1.69   0.090    -.0041785    .0003043
                       |
                   age |   .0049714   .0005786     8.59   0.000     .0038373    .0061055
                female |  -.0685286   .0153908    -4.45   0.000    -.0986939   -.0383633
                unempl |  -.1504654   .0168174    -8.95   0.000     -.183427   -.1175038
                 urban |   -.048071   .0188974    -2.54   0.011    -.0851092   -.0110329
              interest |  -.0261323   .0074271    -3.52   0.000    -.0406891   -.0115755
                  news |  -.0662195   .0057804   -11.46   0.000     -.077549   -.0548901
                       |
                 round |
                    5  |   .1470157   .0235374     6.25   0.000     .1008832    .1931481
                    6  |  -.0843234   .0239895    -3.52   0.000     -.131342   -.0373047
                       |
                 _cons |    3.41868   .0716359    47.72   0.000     3.278277    3.559084
-----------------------+----------------------------------------------------------------
               sigma_u |  .24414612
               sigma_e |  1.0937035
                   rho |  .04746583   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish 

Average marginal effects                                Number of obs = 21,859
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0155547   .0055124     2.82   0.005     .0047505    .0263589
          2  |   .0136176   .0045921     2.97   0.003     .0046172     .022618
          3  |   .0116805   .0037952     3.08   0.002      .004242     .019119
          4  |   .0097434   .0032149     3.03   0.002     .0034423    .0160445
          5  |   .0078062   .0029805     2.62   0.009     .0019646    .0136478
          6  |   .0058691   .0031696     1.85   0.064    -.0003432    .0120814
          7  |    .003932   .0037182     1.06   0.290    -.0033556    .0112196
          8  |   .0019949   .0044967     0.44   0.657    -.0068184    .0108082
          9  |   .0000578   .0054065     0.01   0.991    -.0105388    .0106543
         10  |  -.0018793   .0063919    -0.29   0.769    -.0144072    .0106486
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Bad Living Conditions", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_condlive, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. xtreg condecon c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     21,667
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0209                                         min =        219
     Between = 0.0000                                         avg =    1,083.3
     Overall = 0.0175                                         max =      2,332

                                                Wald chi2(11)     =     461.41
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
              condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0257264   .0057337     4.49   0.000     .0144886    .0369643
               educlvl |   .0137694   .0080451     1.71   0.087    -.0019988    .0295376
                       |
c.lnfdi_jobs#c.educlvl |  -.0023577   .0011862    -1.99   0.047    -.0046826   -.0000328
                       |
                   age |   .0032796   .0005996     5.47   0.000     .0021043    .0044548
                female |   -.034709    .015903    -2.18   0.029    -.0658784   -.0035397
                unempl |  -.0711047   .0173989    -4.09   0.000    -.1052059   -.0370036
                 urban |   .0545116   .0195693     2.79   0.005     .0161565    .0928668
              interest |  -.0455075    .007686    -5.92   0.000    -.0605718   -.0304432
                  news |   -.015779   .0059705    -2.64   0.008     -.027481    -.004077
                       |
                 round |
                    5  |   .2706319   .0243901    11.10   0.000     .2228282    .3184357
                    6  |  -.0296407   .0248403    -1.19   0.233    -.0783268    .0190454
                       |
                 _cons |   3.202355   .1082679    29.58   0.000     2.990154    3.414556
-----------------------+----------------------------------------------------------------
               sigma_u |  .43366607
               sigma_e |  1.1251527
                   rho |  .12934095   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish 

Average marginal effects                                Number of obs = 21,667
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0257264   .0057337     4.49   0.000     .0144886    .0369643
          2  |   .0233687   .0047766     4.89   0.000     .0140066    .0327308
          3  |    .021011   .0039461     5.32   0.000     .0132768    .0287452
          4  |   .0186532   .0033378     5.59   0.000     .0121112    .0251952
          5  |   .0162955   .0030862     5.28   0.000     .0102467    .0223443
          6  |   .0139378   .0032745     4.26   0.000     .0075199    .0203556
          7  |     .01158   .0038385     3.02   0.003     .0040567    .0191033
          8  |   .0092223   .0046433     1.99   0.047     .0001216     .018323
          9  |   .0068646   .0055857     1.23   0.219    -.0040832    .0178124
         10  |   .0045068   .0066071     0.68   0.495    -.0084429    .0174566
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Bad Economic Situation", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_condecon, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. 
. *** FIGURE 2: Distrust in Electoral Commission
. xtreg trustelec c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,583
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0175                                         min =        197
     Between = 0.1990                                         avg =      978.1
     Overall = 0.0217                                         max =      2,258

                                                Wald chi2(11)     =     336.39
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0097377   .0027323     3.56   0.000     .0043825    .0150929
               educlvl |   .0189582   .0036644     5.17   0.000     .0117762    .0261403
                       |
c.lnfdi_jobs#c.educlvl |  -.0017944   .0005684    -3.16   0.002    -.0029084   -.0006804
                       |
                   age |  -.0017365   .0002768    -6.27   0.000     -.002279    -.001194
                female |  -.0082069   .0072625    -1.13   0.258    -.0224411    .0060272
                unempl |  -.0066345   .0080427    -0.82   0.409    -.0223979    .0091289
                 urban |   .0522606   .0090092     5.80   0.000     .0346029    .0699182
              interest |  -.0307532   .0034905    -8.81   0.000    -.0375944    -.023912
                  news |   .0071711   .0027387     2.62   0.009     .0018033    .0125389
                       |
                 round |
                    5  |   .0319239   .0109021     2.93   0.003     .0105561    .0532916
                    6  |  -.0361977   .0109388    -3.31   0.001    -.0576374    -.014758
                       |
                 _cons |   .5458911   .0279087    19.56   0.000     .4911911     .600591
-----------------------+----------------------------------------------------------------
               sigma_u |  .07553513
               sigma_e |  .47701053
                   rho |  .02446168   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish 

Average marginal effects                                Number of obs = 18,583
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0097377   .0027323     3.56   0.000     .0043825    .0150929
          2  |   .0079433    .002262     3.51   0.000     .0035099    .0123767
          3  |    .006149   .0018476     3.33   0.001     .0025276    .0097703
          4  |   .0043546   .0015353     2.84   0.005     .0013455    .0073637
          5  |   .0025602   .0013952     1.83   0.067    -.0001744    .0052947
          6  |   .0007658   .0014772     0.52   0.604    -.0021295    .0036611
          7  |  -.0010286   .0017504    -0.59   0.557    -.0044592    .0024021
          8  |   -.002823   .0021428    -1.32   0.188    -.0070228    .0013769
          9  |  -.0046173   .0026011    -1.78   0.076    -.0097153    .0004806
         10  |  -.0064117    .003096    -2.07   0.038    -.0124798   -.0003436
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Distrust in  Electoral Commission", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_trustelec, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. xtreg trustlead c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,859
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0214                                         min =        211
     Between = 0.0091                                         avg =    1,043.0
     Overall = 0.0203                                         max =      2,294

                                                Wald chi2(11)     =     453.38
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0063255    .002429     2.60   0.009     .0015646    .0110863
               educlvl |   .0153434   .0034325     4.47   0.000     .0086159    .0220709
                       |
c.lnfdi_jobs#c.educlvl |   -.000845   .0005041    -1.68   0.094    -.0018331     .000143
                       |
                   age |  -.0020464   .0002573    -7.95   0.000    -.0025508   -.0015421
                female |  -.0011829   .0068309    -0.17   0.863    -.0145712    .0122054
                unempl |  -.0008925   .0074545    -0.12   0.905    -.0155031    .0137181
                 urban |   .0507245   .0084438     6.01   0.000      .034175    .0672741
              interest |  -.0229246   .0033128    -6.92   0.000    -.0294175   -.0164317
                  news |   .0107462   .0025639     4.19   0.000     .0057211    .0157713
                       |
                 round |
                    5  |   .0509599   .0103856     4.91   0.000     .0306046    .0713153
                    6  |  -.0364752   .0105845    -3.45   0.001    -.0572204   -.0157301
                       |
                 _cons |   .4190262   .0297561    14.08   0.000     .3607054     .477347
-----------------------+----------------------------------------------------------------
               sigma_u |  .09648029
               sigma_e |  .47400348
                   rho |  .03978177   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish 

Average marginal effects                                Number of obs = 20,859
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0063255    .002429     2.60   0.009     .0015646    .0110863
          2  |   .0054804   .0020233     2.71   0.007     .0015148    .0094461
          3  |   .0046354   .0016721     2.77   0.006     .0013582    .0079126
          4  |   .0037904   .0014164     2.68   0.007     .0010143    .0065664
          5  |   .0029453   .0013133     2.24   0.025     .0003714    .0055192
          6  |   .0021003    .001397     1.50   0.133    -.0006377    .0048383
          7  |   .0012552   .0016391     0.77   0.444    -.0019573    .0044678
          8  |   .0004102   .0019824     0.21   0.836    -.0034753    .0042957
          9  |  -.0004348   .0023836    -0.18   0.855    -.0051067     .004237
         10  |  -.0012799   .0028181    -0.45   0.650    -.0068033    .0042435
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Distrust in Country's Leader", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_trustlead, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. xtreg trustpart c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,536
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0269                                         min =        206
     Between = 0.0496                                         avg =    1,028.2
     Overall = 0.0295                                         max =      2,290

                                                Wald chi2(11)     =     541.37
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0091813    .002503     3.67   0.000     .0042755    .0140872
               educlvl |   .0201943   .0035446     5.70   0.000      .013247    .0271416
                       |
c.lnfdi_jobs#c.educlvl |  -.0011215   .0005197    -2.16   0.031    -.0021401   -.0001028
                       |
                   age |   -.001848    .000261    -7.08   0.000    -.0023595   -.0013365
                female |  -.0122745   .0068856    -1.78   0.075      -.02577    .0012211
                unempl |   -.014705   .0075401    -1.95   0.051    -.0294834    .0000734
                 urban |   .0452122   .0086042     5.25   0.000     .0283482    .0620762
              interest |  -.0330453    .003335    -9.91   0.000    -.0395817   -.0265088
                  news |   .0155354   .0025755     6.03   0.000     .0104875    .0205833
                       |
                 round |
                    5  |   .0197332   .0102578     1.92   0.054    -.0003716    .0398381
                    6  |  -.0519232   .0105303    -4.93   0.000    -.0725622   -.0312842
                       |
                 _cons |   .6002221   .0303931    19.75   0.000     .5406527    .6597914
-----------------------+----------------------------------------------------------------
               sigma_u |  .09623867
               sigma_e |   .4629095
                   rho |  .04143146   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish 

Average marginal effects                                Number of obs = 19,536
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0091813    .002503     3.67   0.000     .0042755    .0140872
          2  |   .0080599   .0020806     3.87   0.000      .003982    .0121377
          3  |   .0069384   .0017125     4.05   0.000      .003582    .0102948
          4  |    .005817   .0014411     4.04   0.000     .0029925    .0086414
          5  |   .0046955    .001327     3.54   0.000     .0020946    .0072964
          6  |    .003574   .0014091     2.54   0.011     .0008123    .0063357
          7  |   .0024526   .0016584     1.48   0.139    -.0007978    .0057029
          8  |   .0013311   .0020137     0.66   0.509    -.0026157    .0052779
          9  |   .0002096    .002429     0.09   0.931    -.0045512    .0049705
         10  |  -.0009118   .0028785    -0.32   0.751    -.0065535    .0047299
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Distrust in Ruling Coalition", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_trustpart, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. 
. *** FIGURE 3: Opinion: Country Is Not a Democracy
. xtreg demoqua c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,139
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0186                                         min =        205
     Between = 0.1980                                         avg =    1,007.0
     Overall = 0.0263                                         max =      2,130

                                                Wald chi2(11)     =     385.36
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0182777   .0047273     3.87   0.000     .0090123     .027543
               educlvl |   .0539733    .006476     8.33   0.000     .0412806    .0666661
                       |
c.lnfdi_jobs#c.educlvl |  -.0041256    .000961    -4.29   0.000    -.0060091   -.0022422
                       |
                   age |  -.0004586   .0004815    -0.95   0.341    -.0014022    .0004851
                female |  -.0645194    .012599    -5.12   0.000     -.089213   -.0398258
                unempl |   -.031379   .0138434    -2.27   0.023    -.0585116   -.0042464
                 urban |   .0684748   .0156761     4.37   0.000     .0377503    .0991993
              interest |   -.050375   .0061539    -8.19   0.000    -.0624364   -.0383136
                  news |   .0092748      .0047     1.97   0.048      .000063    .0184866
                       |
                 round |
                    5  |   .1568107   .0192935     8.13   0.000     .1189961    .1946253
                    6  |   .0677677   .0198699     3.41   0.001     .0288234    .1067121
                       |
                 _cons |   2.403199   .0574776    41.81   0.000     2.290545    2.515853
-----------------------+----------------------------------------------------------------
               sigma_u |  .18787816
               sigma_e |  .86112649
                   rho |  .04543836   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish

Average marginal effects                                Number of obs = 20,139
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0182777   .0047273     3.87   0.000     .0090123     .027543
          2  |   .0141521   .0039367     3.59   0.000     .0064362    .0218679
          3  |   .0100264   .0032396     3.09   0.002     .0036769     .016376
          4  |   .0059008   .0027091     2.18   0.029      .000591    .0112106
          5  |   .0017752   .0024557     0.72   0.470    -.0030379    .0065883
          6  |  -.0023504   .0025629    -0.92   0.359    -.0073736    .0026728
          7  |   -.006476   .0029922    -2.16   0.030    -.0123407   -.0006114
          8  |  -.0106016   .0036311    -2.92   0.004    -.0177185   -.0034847
          9  |  -.0147272   .0043891    -3.36   0.001    -.0233297   -.0061248
         10  |  -.0188529   .0052144    -3.62   0.000    -.0290728   -.0086329
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Opinion: Country Is Not a Democracy", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_demoqua, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. xtreg demosat c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,486
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0216                                         min =        208
     Between = 0.0328                                         avg =    1,024.3
     Overall = 0.0238                                         max =      2,161

                                                Wald chi2(11)     =     453.15
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .015368   .0050823     3.02   0.002     .0054068    .0253292
               educlvl |   .0461267   .0070132     6.58   0.000      .032381    .0598724
                       |
c.lnfdi_jobs#c.educlvl |  -.0032212   .0010383    -3.10   0.002    -.0052563   -.0011862
                       |
                   age |  -.0008282   .0005226    -1.58   0.113    -.0018525    .0001961
                female |  -.0722404   .0137055    -5.27   0.000    -.0991027   -.0453781
                unempl |  -.0113564   .0150483    -0.75   0.450    -.0408504    .0181377
                 urban |   .1014529   .0170336     5.96   0.000     .0680677    .1348381
              interest |  -.0869897   .0066793   -13.02   0.000     -.100081   -.0738984
                  news |   .0206825    .005128     4.03   0.000     .0106318    .0307331
                       |
                 round |
                    5  |   .0658503   .0210179     3.13   0.002      .024656    .1070446
                    6  |  -.0550626   .0216036    -2.55   0.011    -.0974049   -.0127203
                       |
                 _cons |   1.748807   .0733794    23.83   0.000     1.604986    1.892628
-----------------------+----------------------------------------------------------------
               sigma_u |  .26800671
               sigma_e |  .94534811
                   rho |  .07439338   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish

Average marginal effects                                Number of obs = 20,486
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |    .015368   .0050823     3.02   0.002     .0054068    .0253292
          2  |   .0121468   .0042331     2.87   0.004     .0038501    .0204434
          3  |   .0089255   .0034877     2.56   0.010     .0020898    .0157612
          4  |   .0057043   .0029266     1.95   0.051    -.0000317    .0114403
          5  |   .0024831   .0026687     0.93   0.352    -.0027476    .0077137
          6  |  -.0007382   .0027992    -0.26   0.792    -.0062245    .0047482
          7  |  -.0039594   .0032719    -1.21   0.226    -.0103722    .0024533
          8  |  -.0071807   .0039662    -1.81   0.070    -.0149543    .0005929
          9  |  -.0104019   .0047867    -2.17   0.030    -.0197836   -.0010202
         10  |  -.0136231   .0056789    -2.40   0.016    -.0247536   -.0024926
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Dissatisfaction with Quality of Democracy", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_demosat, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. xtreg demosup c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     17,634
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0022                                         min =        190
     Between = 0.1183                                         avg =      881.7
     Overall = 0.0010                                         max =      1,990

                                                Wald chi2(11)     =      38.52
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0001

----------------------------------------------------------------------------------------
               demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0071492   .0021238     3.37   0.001     .0029865    .0113118
               educlvl |   .0118591   .0028754     4.12   0.000     .0062234    .0174948
                       |
c.lnfdi_jobs#c.educlvl |  -.0013251   .0004271    -3.10   0.002    -.0021621   -.0004881
                       |
                   age |   .0007002    .000215     3.26   0.001     .0002789    .0011215
                female |   .0006325   .0056377     0.11   0.911    -.0104171    .0116821
                unempl |   .0028936   .0061919     0.47   0.640    -.0092422    .0150294
                 urban |   .0068123   .0069703     0.98   0.328    -.0068492    .0204738
              interest |   .0006507   .0027789     0.23   0.815    -.0047959    .0060973
                  news |   .0018242   .0020981     0.87   0.385     -.002288    .0059364
                       |
                 round |
                    5  |  -.0112223   .0086324    -1.30   0.194    -.0281414    .0056968
                    6  |  -.0148911   .0088882    -1.68   0.094    -.0323117    .0025295
                       |
                 _cons |   .7361241    .023846    30.87   0.000     .6893869    .7828614
-----------------------+----------------------------------------------------------------
               sigma_u |  .07216385
               sigma_e |  .36011564
                   rho |  .03860617   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_jobs) at(educlvl=(0(1)9)) vsquish

Average marginal effects                                Number of obs = 17,634
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_jobs
1._at:  educlvl = 0
2._at:  educlvl = 1
3._at:  educlvl = 2
4._at:  educlvl = 3
5._at:  educlvl = 4
6._at:  educlvl = 5
7._at:  educlvl = 6
8._at:  educlvl = 7
9._at:  educlvl = 8
10._at: educlvl = 9

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_jobs   |
         _at |
          1  |   .0071492   .0021238     3.37   0.001     .0029865    .0113118
          2  |    .005824   .0017711     3.29   0.001     .0023528    .0092953
          3  |   .0044989   .0014586     3.08   0.002     .0016401    .0073577
          4  |   .0031738   .0012178     2.61   0.009     .0007869    .0055607
          5  |   .0018487    .001097     1.69   0.092    -.0003015    .0039988
          6  |   .0005236   .0011352     0.46   0.645    -.0017013    .0027485
          7  |  -.0008016   .0013185    -0.61   0.543    -.0033858    .0017827
          8  |  -.0021267   .0015979    -1.33   0.183    -.0052584    .0010051
          9  |  -.0034518    .001932    -1.79   0.074    -.0072384    .0003348
         10  |  -.0047769   .0022971    -2.08   0.038    -.0092792   -.0002747
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Support for Democracy", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(0 9)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(afb_demosup, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl if sample==1, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(0 `" "no formal" "education" "' 3 `" "primary" "completed" "' ///
>                         6 `" "secondary" "completed" "' 9 `" "teritary" "completed" "', labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. 
. *** FIGURE A-3: Marginal Effects Plots - All Models
. graph combine afb_problem afb_condlive afb_condecon afb_trustelec afb_trustlead ///
>         afb_trustpart afb_demoqua afb_demosat afb_demosup, cols(3) iscale(.4) ///
>         xsize(5) ysize(6) name(afb_alleffects, replace)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)

. 
. *** TABLE 2: Substantial Effect Sizes (and TABLE A-11: All Models)
. * Self-Reported Economic Problems
. xtprobit problem c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -14995.438  
Iteration 1:   log likelihood = -14869.996  
Iteration 2:   log likelihood = -14869.984  
Iteration 3:   log likelihood = -14869.984  

Fitting full model:

rho =  0.0     log likelihood = -14869.984
rho =  0.1     log likelihood = -14397.708
rho =  0.2     log likelihood =   -14452.2

Iteration 0:   log likelihood = -14385.026  
Iteration 1:   log likelihood = -14346.585  
Iteration 2:   log likelihood = -14346.579  
Iteration 3:   log likelihood = -14346.579  

Random-effects probit regression                    Number of obs    =  21,644
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     218
                                                                 avg = 1,082.2
                                                                 max =   2,334

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  232.96
Log likelihood = -14346.579                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0356572   .0065493     5.44   0.000     .0228208    .0484935
               educlvl |    .063722   .0091895     6.93   0.000     .0457108    .0817332
                       |
c.lnfdi_jobs#c.educlvl |  -.0082417   .0013572    -6.07   0.000    -.0109017   -.0055816
                       |
                   age |  -.0040879   .0006813    -6.00   0.000    -.0054231   -.0027526
                female |   .0077719   .0180642     0.43   0.667    -.0276333     .043177
                unempl |   -.075142    .019769    -3.80   0.000    -.1138885   -.0363956
                 urban |   .0958589   .0222114     4.32   0.000     .0523253    .1393925
              interest |  -.0196636   .0087148    -2.26   0.024    -.0367442    -.002583
                  news |   .0131094   .0067791     1.93   0.053    -.0001773    .0263962
                       |
                 round |
                    5  |  -.1311325    .027481    -4.77   0.000    -.1849942   -.0772708
                    6  |  -.1585194   .0280467    -5.65   0.000    -.2134899   -.1035489
                       |
                 _cons |  -.0277421   .0914861    -0.30   0.762    -.2070515    .1515673
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.231733   .3237187                      -2.86621   -1.597256
-----------------------+----------------------------------------------------------------
               sigma_u |   .3276313   .0530302                       .238567    .4499459
                   rho |   .0969368   .0283384                      .0538494    .1683655
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 1046.81                Prob >= chibar2 = 0.000

. margins, post ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 21,644
Model VCE: OIM

Expression: Pr(problem=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4688151   .0286675    16.35   0.000     .4126277    .5250024
          2  |    .512437   .0288392    17.77   0.000     .4559133    .5689607
          3  |   .5401422   .0287773    18.77   0.000     .4837398    .5965446
          4  |   .5031482   .0287331    17.51   0.000     .4468323     .559464
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 21,644
Model VCE: OIM

Expression: Pr(problem=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.57868 (mean)
       female     =  .499307 (mean)
       unempl     = .3158381 (mean)
       urban      = .7122066 (mean)
       interest   =  1.64452 (mean)
       news       = 1.535437 (mean)
       4.round    = .1779708 (mean)
       5.round    =  .406995 (mean)
       6.round    = .4150342 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .4688151  _b[1bn._at]
          2  |    .512437  _b[2._at]
          3  |   .5401422  _b[3._at]
          4  |   .5031482  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0436219   .0102503     4.26   0.000     .0235318    .0637121
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =   18.11
         Prob > chi2 =    0.0000

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   -.036994   .0104848    -3.53   0.000    -.0575438   -.0164442
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =   12.45
         Prob > chi2 =    0.0004

. xtoprobit condlive c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -32442.445  
Iteration 1:   log likelihood = -31835.215  
Iteration 2:   log likelihood = -31835.077  
Iteration 3:   log likelihood = -31835.077  

Refining starting values:

Grid node 0:   log likelihood = -31390.028

Fitting full model:

Iteration 0:   log likelihood = -31390.028  (not concave)
Iteration 1:   log likelihood = -31384.254  (not concave)
Iteration 2:   log likelihood = -31378.461  (not concave)
Iteration 3:   log likelihood = -31373.039  (not concave)
Iteration 4:   log likelihood = -31368.064  (not concave)
Iteration 5:   log likelihood = -31365.994  (not concave)
Iteration 6:   log likelihood = -31364.476  
Iteration 7:   log likelihood = -31338.227  
Iteration 8:   log likelihood =  -31338.09  
Iteration 9:   log likelihood = -31338.089  

Random-effects ordered probit regression            Number of obs    =  21,859
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     219
                                                                 avg = 1,093.0
                                                                 max =   2,337

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  960.14
Log likelihood = -31338.089                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
              condlive | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0167539   .0053313     3.14   0.002     .0063047     .027203
               educlvl |  -.0380796   .0074939    -5.08   0.000    -.0527674   -.0233919
                       |
c.lnfdi_jobs#c.educlvl |  -.0022978   .0011062    -2.08   0.038     -.004466   -.0001296
                       |
                   age |   .0047344   .0005601     8.45   0.000     .0036366    .0058322
                female |  -.0649059    .014866    -4.37   0.000    -.0940427   -.0357691
                unempl |  -.1426788   .0162526    -8.78   0.000    -.1745333   -.1108243
                 urban |  -.0469033   .0182525    -2.57   0.010    -.0826776   -.0111291
              interest |   -.025526   .0071885    -3.55   0.000    -.0396152   -.0114368
                  news |   -.063226   .0055885   -11.31   0.000    -.0741792   -.0522728
                       |
                 round |
                    5  |   .1346365   .0227277     5.92   0.000      .090091    .1791821
                    6  |  -.0890107   .0231585    -3.84   0.000    -.1344004   -.0436209
-----------------------+----------------------------------------------------------------
                 /cut1 |  -2.005588   .0689784                     -2.140783   -1.870393
                 /cut2 |  -.6495524   .0676301                     -.7821049   -.5169999
                 /cut3 |   .0105364   .0675287                     -.1218174    .1428902
                 /cut4 |   .8586738   .0677238                      .7259376    .9914101
-----------------------+----------------------------------------------------------------
             /sigma2_u |   .0506501   .0163571                      .0268962    .0953829
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 993.98       Prob >= chibar2 = 0.0000

. margins, post predict(outcome(5)) ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 21,859
Model VCE: OIM

Expression: Pr(5.condlive), predict(outcome(5))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .1723437   .0133935    12.87   0.000     .1460929    .1985945
          2  |   .1910924    .014347    13.32   0.000     .1629729     .219212
          3  |   .1294355    .011221    11.54   0.000     .1074427    .1514283
          4  |   .1322305   .0113083    11.69   0.000     .1100668    .1543943
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 21,859
Model VCE: OIM

Expression: Pr(5.condlive), predict(outcome(5))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.59664 (mean)
       female     = .4987419 (mean)
       unempl     = .3163914 (mean)
       urban      = .7125669 (mean)
       interest   = 1.645409 (mean)
       news       = 1.533968 (mean)
       4.round    =  .177547 (mean)
       5.round    = .4057825 (mean)
       6.round    = .4166705 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .1723437  _b[1bn._at]
          2  |   .1910924  _b[2._at]
          3  |   .1294355  _b[3._at]
          4  |   .1322305  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0187487   .0057312     3.27   0.001     .0075158    .0299817
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =   10.70
         Prob > chi2 =    0.0011

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |    .002795   .0047171     0.59   0.553    -.0064504    .0120404
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    0.35
         Prob > chi2 =    0.5535

. xtoprobit condecon c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -32159.466  
Iteration 1:   log likelihood = -31912.682  
Iteration 2:   log likelihood = -31912.668  
Iteration 3:   log likelihood = -31912.668  

Refining starting values:

Grid node 0:   log likelihood =   -30515.4

Fitting full model:

Iteration 0:   log likelihood =   -30515.4  (not concave)
Iteration 1:   log likelihood = -30510.367  (not concave)
Iteration 2:   log likelihood = -30505.332  (not concave)
Iteration 3:   log likelihood = -30501.121  (not concave)
Iteration 4:   log likelihood = -30497.648  (not concave)
Iteration 5:   log likelihood = -30496.322  
Iteration 6:   log likelihood = -30371.037  
Iteration 7:   log likelihood = -30369.811  
Iteration 8:   log likelihood = -30369.693  
Iteration 9:   log likelihood = -30369.692  

Random-effects ordered probit regression            Number of obs    =  21,667
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     219
                                                                 avg = 1,083.3
                                                                 max =   2,332

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  455.16
Log likelihood = -30369.692                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
              condecon | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0264143   .0054528     4.84   0.000     .0157271    .0371016
               educlvl |   .0159775   .0076442     2.09   0.037     .0009951    .0309599
                       |
c.lnfdi_jobs#c.educlvl |   -.002558   .0011306    -2.26   0.024    -.0047739   -.0003422
                       |
                   age |   .0030701   .0005714     5.37   0.000     .0019502      .00419
                female |  -.0329792   .0151214    -2.18   0.029    -.0626166   -.0033417
                unempl |  -.0653754   .0165293    -3.96   0.000    -.0977721   -.0329786
                 urban |   .0562376   .0186216     3.02   0.003     .0197399    .0927354
              interest |  -.0431722   .0073335    -5.89   0.000    -.0575456   -.0287989
                  news |  -.0148089   .0056827    -2.61   0.009    -.0259469    -.003671
                       |
                 round |
                    5  |   .2419917   .0232195    10.42   0.000     .1964823     .287501
                    6  |  -.0476169   .0235499    -2.02   0.043    -.0937738   -.0014599
-----------------------+----------------------------------------------------------------
                 /cut1 |  -1.675827    .102189                     -1.876114    -1.47554
                 /cut2 |  -.3523513   .1012625                      -.550822   -.1538805
                 /cut3 |   .1555041   .1012366                     -.0429159    .3539241
                 /cut4 |   .9776878   .1013842                      .7789784    1.176397
-----------------------+----------------------------------------------------------------
             /sigma2_u |   .1625732   .0517821                      .0870819    .3035079
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 3085.95      Prob >= chibar2 = 0.0000

. margins, post predict(outcome(5)) ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 21,667
Model VCE: OIM

Expression: Pr(5.condecon), predict(outcome(5))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .2297485   .0263327     8.72   0.000     .1781374    .2813596
          2  |   .2671649   .0284354     9.40   0.000     .2114326    .3228971
          3  |   .2417552    .027097     8.92   0.000     .1886461    .2948643
          4  |   .2598878    .028001     9.28   0.000     .2050068    .3147688
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 21,667
Model VCE: OIM

Expression: Pr(5.condecon), predict(outcome(5))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.52495 (mean)
       female     = .4971154 (mean)
       unempl     = .3143029 (mean)
       urban      = .7138967 (mean)
       interest   = 1.651636 (mean)
       news       = 1.542992 (mean)
       4.round    = .1767204 (mean)
       5.round    = .4054091 (mean)
       6.round    = .4178705 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .2297485  _b[1bn._at]
          2  |   .2671649  _b[2._at]
          3  |   .2417552  _b[3._at]
          4  |   .2598878  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0374163   .0069543     5.38   0.000     .0237862    .0510465
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =   28.95
         Prob > chi2 =    0.0000

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0181326   .0069103     2.62   0.009     .0045887    .0316765
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    6.89
         Prob > chi2 =    0.0087

. * Distrust in Electoral Commission
. xtprobit trustelec c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -12743.597  
Iteration 1:   log likelihood =  -12519.49  
Iteration 2:   log likelihood = -12519.425  
Iteration 3:   log likelihood = -12519.425  

Fitting full model:

rho =  0.0     log likelihood = -12519.425
rho =  0.1     log likelihood = -12083.922
rho =  0.2     log likelihood = -12118.673

Iteration 0:   log likelihood = -12066.828  
Iteration 1:   log likelihood = -12046.868  
Iteration 2:   log likelihood = -12046.852  
Iteration 3:   log likelihood = -12046.852  

Random-effects probit regression                     Number of obs    = 18,583
Group variable: cname                                Number of groups =     19

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    197
                                                                  avg =  978.1
                                                                  max =  2,258

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(11)    = 322.70
Log likelihood = -12046.852                          Prob > chi2      = 0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0258788   .0073287     3.53   0.000     .0115148    .0402427
               educlvl |   .0496119   .0098623     5.03   0.000     .0302821    .0689417
                       |
c.lnfdi_jobs#c.educlvl |  -.0047635   .0015313    -3.11   0.002    -.0077648   -.0017622
                       |
                   age |  -.0046147   .0007464    -6.18   0.000    -.0060777   -.0031517
                female |  -.0218504   .0196296    -1.11   0.266    -.0603237    .0166228
                unempl |  -.0184464   .0217351    -0.85   0.396    -.0610464    .0241535
                 urban |   .1386468   .0242331     5.72   0.000     .0911507    .1861429
              interest |  -.0825382   .0094489    -8.74   0.000    -.1010576   -.0640187
                  news |   .0199036   .0074229     2.68   0.007     .0053549    .0344524
                       |
                 round |
                    5  |   .0837043   .0295771     2.83   0.005     .0257342    .1416744
                    6  |  -.0976622   .0296277    -3.30   0.001    -.1557314    -.039593
                       |
                 _cons |   .1292632   .0896826     1.44   0.149    -.0465115     .305038
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.444932   .3337316                     -3.099034    -1.79083
-----------------------+----------------------------------------------------------------
               sigma_u |   .2945031   .0491425                      .2123506    .4084381
                   rho |     .07981   .0245094                      .0431471     .142971
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 945.15                 Prob >= chibar2 = 0.000

. margins, post ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 18,583
Model VCE: OIM

Expression: Pr(trustelec=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5273395    .027017    19.52   0.000     .4743872    .5802918
          2  |   .5646059   .0270531    20.87   0.000     .5115828     .617629
          3  |   .5894035   .0265591    22.19   0.000     .5373486    .6414584
          4  |   .5800569   .0266334    21.78   0.000     .5278564    .6322574
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 18,583
Model VCE: OIM

Expression: Pr(trustelec=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.06522 (mean)
       female     = .4911478 (mean)
       unempl     = .2989829 (mean)
       urban      = .7104881 (mean)
       interest   = 1.660496 (mean)
       news       = 1.601195 (mean)
       4.round    = .1948555 (mean)
       5.round    = .3686703 (mean)
       6.round    = .4364742 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .5273395  _b[1bn._at]
          2  |   .5646059  _b[2._at]
          3  |   .5894035  _b[3._at]
          4  |   .5800569  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0372664   .0112824     3.30   0.001     .0151532    .0593795
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =   10.91
         Prob > chi2 =    0.0010

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0093466   .0113102    -0.83   0.409    -.0315143     .012821
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    0.68
         Prob > chi2 =    0.4086

. xtprobit trustlead c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -14343.917  
Iteration 1:   log likelihood = -14087.024  
Iteration 2:   log likelihood = -14086.896  
Iteration 3:   log likelihood = -14086.896  

Fitting full model:

rho =  0.0     log likelihood = -14086.896
rho =  0.1     log likelihood = -13456.036
rho =  0.2     log likelihood = -13512.293

Iteration 0:   log likelihood = -13427.411  
Iteration 1:   log likelihood = -13388.745  
Iteration 2:   log likelihood = -13388.743  
Iteration 3:   log likelihood = -13388.743  

Random-effects probit regression                    Number of obs    =  20,859
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     211
                                                                 avg = 1,043.0
                                                                 max =   2,294

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  440.00
Log likelihood = -13388.743                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
             trustlead | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0174085   .0066739     2.61   0.009      .004328     .030489
               educlvl |   .0401229   .0093574     4.29   0.000     .0217827    .0584631
                       |
c.lnfdi_jobs#c.educlvl |  -.0022394   .0013748    -1.63   0.103    -.0049339    .0004551
                       |
                   age |  -.0057456    .000708    -8.12   0.000    -.0071332    -.004358
                female |  -.0041467   .0186594    -0.22   0.824    -.0407184     .032425
                unempl |   -.003801   .0203292    -0.19   0.852    -.0436455    .0360436
                 urban |   .1403341   .0231335     6.07   0.000     .0949932     .185675
              interest |  -.0634645   .0090588    -7.01   0.000    -.0812195   -.0457096
                  news |   .0306141   .0070066     4.37   0.000     .0168815    .0443467
                       |
                 round |
                    5  |   .1372572   .0285601     4.81   0.000     .0812805     .193234
                    6  |  -.0952956   .0291736    -3.27   0.001    -.1524749   -.0381163
                       |
                 _cons |  -.2118935   .0928641    -2.28   0.023    -.3939038   -.0298833
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.214996   .3216899                     -2.845497   -1.584495
-----------------------+----------------------------------------------------------------
               sigma_u |   .3303845   .0531407                      .2410506    .4528258
                   rho |   .0984119   .0285426                      .0549146    .1701598
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 1396.31                Prob >= chibar2 = 0.000

. margins, post ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 20,859
Model VCE: OIM

Expression: Pr(trustlead=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |    .405295   .0282261    14.36   0.000     .3499728    .4606172
          2  |   .4340307   .0287237    15.11   0.000     .3777333     .490328
          3  |   .4611657   .0290281    15.89   0.000     .4042716    .5180597
          4  |   .4685618    .028894    16.22   0.000     .4119306    .5251929
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 20,859
Model VCE: OIM

Expression: Pr(trustlead=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.51877 (mean)
       female     = .4957572 (mean)
       unempl     = .3160746 (mean)
       urban      = .7177238 (mean)
       interest   = 1.656983 (mean)
       news       = 1.530179 (mean)
       4.round    = .1805935 (mean)
       5.round    = .3915336 (mean)
       6.round    = .4278729 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |    .405295  _b[1bn._at]
          2  |   .4340307  _b[2._at]
          3  |   .4611657  _b[3._at]
          4  |   .4685618  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0287357     .01026     2.80   0.005     .0086265    .0488449
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    7.84
         Prob > chi2 =    0.0051

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0073961   .0106772     0.69   0.488    -.0135307     .028323
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    0.48
         Prob > chi2 =    0.4885

. xtprobit trustpart c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -13210.329  
Iteration 1:   log likelihood = -12861.322  
Iteration 2:   log likelihood = -12861.056  
Iteration 3:   log likelihood = -12861.056  

Fitting full model:

rho =  0.0     log likelihood = -12861.056
rho =  0.1     log likelihood = -12158.605
rho =  0.2     log likelihood = -12226.525

Iteration 0:   log likelihood =  -12147.66  
Iteration 1:   log likelihood = -12099.434  
Iteration 2:   log likelihood =   -12099.4  
Iteration 3:   log likelihood =   -12099.4  

Random-effects probit regression                    Number of obs    =  19,536
Group variable: cname                               Number of groups =      19

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     206
                                                                 avg = 1,028.2
                                                                 max =   2,290

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  513.96
Log likelihood = -12099.4                           Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
             trustpart | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0251153   .0070779     3.55   0.000     .0112428    .0389877
               educlvl |   .0548366   .0101037     5.43   0.000     .0350338    .0746395
                       |
c.lnfdi_jobs#c.educlvl |  -.0028134   .0014888    -1.89   0.059    -.0057313    .0001046
                       |
                   age |  -.0050685   .0007382    -6.87   0.000    -.0065155   -.0036216
                female |  -.0343838   .0195452    -1.76   0.079    -.0726917    .0039241
                unempl |  -.0414162   .0214002    -1.94   0.053    -.0833598    .0005273
                 urban |   .1230136   .0242098     5.08   0.000     .0755632     .170464
              interest |  -.0935668   .0094856    -9.86   0.000    -.1121582   -.0749754
                  news |   .0436232   .0073319     5.95   0.000      .029253    .0579935
                       |
                 round |
                    5  |    .052063   .0290851     1.79   0.073    -.0049428    .1090687
                    6  |  -.1470489   .0297802    -4.94   0.000     -.205417   -.0886808
                       |
                 _cons |   .2812132   .1038769     2.71   0.007     .0776183    .4848081
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -1.975658   .3294007                     -2.621272   -1.330045
-----------------------+----------------------------------------------------------------
               sigma_u |   .3723842   .0613318                      .2696485     .514262
                   rho |   .1217824   .0352299                      .0677819     .209152
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 1523.31                Prob >= chibar2 = 0.000

. margins, post ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 19,536
Model VCE: OIM

Expression: Pr(trustpart=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5711565   .0324245    17.61   0.000     .5076057    .6347073
          2  |   .6135914    .031732    19.34   0.000     .5513978    .6757849
          3  |    .645623   .0310202    20.81   0.000     .5848245    .7064214
          4  |    .660853   .0303418    21.78   0.000     .6013841    .7203219
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 19,536
Model VCE: OIM

Expression: Pr(trustpart=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.35944 (mean)
       female     = .4936527 (mean)
       unempl     = .3107596 (mean)
       urban      = .7269656 (mean)
       interest   = 1.667793 (mean)
       news       = 1.539056 (mean)
       4.round    =  .190776 (mean)
       5.round    = .4064803 (mean)
       6.round    = .4027437 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .5711565  _b[1bn._at]
          2  |   .6135914  _b[2._at]
          3  |    .645623  _b[3._at]
          4  |    .660853  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0424349   .0105635     4.02   0.000     .0217307     .063139
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =   16.14
         Prob > chi2 =    0.0001

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |     .01523   .0105608     1.44   0.149    -.0054687    .0359288
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    2.08
         Prob > chi2 =    0.1493

. * Opinion: Country Is Not a Democracy
. xtoprobit demoqua c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood = -26160.975  
Iteration 1:   log likelihood = -25831.947  
Iteration 2:   log likelihood = -25831.912  
Iteration 3:   log likelihood = -25831.912  

Refining starting values:

Grid node 0:   log likelihood = -25029.431

Fitting full model:

Iteration 0:   log likelihood = -25029.431  (not concave)
Iteration 1:   log likelihood = -25020.198  (not concave)
Iteration 2:   log likelihood = -25015.996  (not concave)
Iteration 3:   log likelihood = -25012.434  (not concave)
Iteration 4:   log likelihood = -25009.552  
Iteration 5:   log likelihood = -24949.468  
Iteration 6:   log likelihood = -24928.142  
Iteration 7:   log likelihood = -24927.996  
Iteration 8:   log likelihood = -24927.996  

Random-effects ordered probit regression            Number of obs    =  20,139
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     205
                                                                 avg = 1,007.0
                                                                 max =   2,130

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  381.62
Log likelihood = -24927.996                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .023138   .0059087     3.92   0.000     .0115572    .0347187
               educlvl |   .0674269   .0080936     8.33   0.000     .0515637    .0832901
                       |
c.lnfdi_jobs#c.educlvl |  -.0052394   .0012004    -4.36   0.000    -.0075922   -.0028867
                       |
                   age |  -.0005951   .0006013    -0.99   0.322    -.0017737    .0005835
                female |  -.0816032   .0157243    -5.19   0.000    -.1124222   -.0507842
                unempl |  -.0396457   .0172783    -2.29   0.022    -.0735105   -.0057809
                 urban |   .0846611   .0195801     4.32   0.000     .0462848    .1230375
              interest |  -.0633085   .0076967    -8.23   0.000    -.0783938   -.0482231
                  news |   .0122307   .0058717     2.08   0.037     .0007224     .023739
                       |
                 round |
                    5  |   .1966151   .0241068     8.16   0.000     .1493667    .2438636
                    6  |   .0864559   .0248205     3.48   0.000     .0378086    .1351031
-----------------------+----------------------------------------------------------------
                 /cut1 |   -.999314    .090596                     -1.176879   -.8217491
                 /cut2 |   .1182582   .0903845                      -.058892    .2954085
                 /cut3 |   1.241348   .0906286                       1.06372    1.418977
-----------------------+----------------------------------------------------------------
             /sigma2_u |   .1147278   .0368343                      .0611482    .2152552
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 1807.83      Prob >= chibar2 = 0.0000

. margins, post predict(outcome(4)) ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 20,139
Model VCE: OIM

Expression: Pr(4.demoqua), predict(outcome(4))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .1582387   .0183514     8.62   0.000     .1222707    .1942067
          2  |   .1762565   .0196396     8.97   0.000     .1377635    .2147495
          3  |   .2182422   .0221582     9.85   0.000     .1748129    .2616715
          4  |   .2020433   .0211313     9.56   0.000     .1606266      .24346
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 20,139
Model VCE: OIM

Expression: Pr(4.demoqua), predict(outcome(4))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.27802 (mean)
       female     = .4841353 (mean)
       unempl     = .3065693 (mean)
       urban      = .7242663 (mean)
       interest   = 1.687273 (mean)
       news       = 1.604151 (mean)
       4.round    = .1746363 (mean)
       5.round    = .4149163 (mean)
       6.round    = .4104474 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .1582387  _b[1bn._at]
          2  |   .1762565  _b[2._at]
          3  |   .2182422  _b[3._at]
          4  |   .2020433  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0180178   .0058977     3.06   0.002     .0064584    .0295772
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    9.33
         Prob > chi2 =    0.0023

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0161989   .0065489    -2.47   0.013    -.0290345   -.0033633
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    6.12
         Prob > chi2 =    0.0134

. xtoprobit demosat c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood =  -28841.08  
Iteration 1:   log likelihood = -28529.981  
Iteration 2:   log likelihood = -28529.944  
Iteration 3:   log likelihood = -28529.944  

Refining starting values:

Grid node 0:   log likelihood =  -27598.41

Fitting full model:

Iteration 0:   log likelihood =  -27598.41  (not concave)
Iteration 1:   log likelihood = -27593.814  (not concave)
Iteration 2:   log likelihood = -27589.216  (not concave)
Iteration 3:   log likelihood = -27585.331  (not concave)
Iteration 4:   log likelihood = -27582.082  (not concave)
Iteration 5:   log likelihood = -27579.393  (not concave)
Iteration 6:   log likelihood = -27577.219  
Iteration 7:   log likelihood = -27532.517  
Iteration 8:   log likelihood = -27474.053  
Iteration 9:   log likelihood = -27473.985  
Iteration 10:  log likelihood = -27473.985  

Random-effects ordered probit regression            Number of obs    =  20,486
Group variable: cname                               Number of groups =      20

Random effects u_i ~ Gaussian                       Obs per group:
                                                                 min =     208
                                                                 avg = 1,024.3
                                                                 max =   2,161

Integration method: mvaghermite                     Integration pts. =      12

                                                    Wald chi2(11)    =  442.99
Log likelihood = -27473.985                         Prob > chi2      =  0.0000

----------------------------------------------------------------------------------------
               demosat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0174404   .0056888     3.07   0.002     .0062906    .0285903
               educlvl |    .051695   .0078472     6.59   0.000     .0363146    .0670753
                       |
c.lnfdi_jobs#c.educlvl |  -.0036305    .001161    -3.13   0.002     -.005906   -.0013549
                       |
                   age |  -.0009712   .0005853    -1.66   0.097    -.0021183    .0001759
                female |   -.081706   .0153293    -5.33   0.000    -.1117509    -.051661
                unempl |  -.0118755   .0168229    -0.71   0.480    -.0448479    .0210968
                 urban |   .1116489   .0190666     5.86   0.000      .074279    .1490188
              interest |  -.0969082    .007498   -12.92   0.000     -.111604   -.0822125
                  news |      .0239    .005738     4.17   0.000     .0126538    .0351462
                       |
                 round |
                    5  |   .0736518   .0235412     3.13   0.002     .0275118    .1197918
                    6  |  -.0556039    .024203    -2.30   0.022     -.103041   -.0081668
-----------------------+----------------------------------------------------------------
                 /cut1 |  -1.334088   .0929566                      -1.51628   -1.151897
                 /cut2 |  -.1983547   .0925562                     -.3797615   -.0169478
                 /cut3 |   .7220226   .0926215                      .5404878    .9035573
                 /cut4 |   2.054984   .0939541                      1.870837    2.239131
-----------------------+----------------------------------------------------------------
             /sigma2_u |    .125837   .0402034                      .0672759    .2353732
----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 2111.92      Prob >= chibar2 = 0.0000

. margins, post predict(outcome(3)) ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 20,486
Model VCE: OIM

Expression: Pr(3.demosat), predict(outcome(3))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .2249642   .0198103    11.36   0.000     .1861368    .2637916
          2  |   .2398769   .0202011    11.87   0.000     .2002835    .2794704
          3  |   .2707855   .0205993    13.15   0.000     .2304116    .3111594
          4  |   .2625742   .0204518    12.84   0.000     .2224894     .302659
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 20,486
Model VCE: OIM

Expression: Pr(3.demosat), predict(outcome(3))
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.32705 (mean)
       female     = .4868203 (mean)
       unempl     = .3092844 (mean)
       urban      = .7224446 (mean)
       interest   = 1.680855 (mean)
       news       = 1.590062 (mean)
       4.round    =  .175144 (mean)
       5.round    = .4141365 (mean)
       6.round    = .4107195 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .2249642  _b[1bn._at]
          2  |   .2398769  _b[2._at]
          3  |   .2707855  _b[3._at]
          4  |   .2625742  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0149128   .0057309     2.60   0.009     .0036803    .0261452
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    6.77
         Prob > chi2 =    0.0093

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0082113   .0058358    -1.41   0.159    -.0196492    .0032266
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    1.98
         Prob > chi2 =    0.1594

. xtprobit demosup c.lnfdi_jobs##c.educlvl $C1 $C2 i.round, i(cname)

Fitting comparison model:

Iteration 0:   log likelihood =  -7833.167  
Iteration 1:   log likelihood = -7802.5805  
Iteration 2:   log likelihood = -7802.5317  
Iteration 3:   log likelihood = -7802.5317  

Fitting full model:

rho =  0.0     log likelihood = -7802.5317
rho =  0.1     log likelihood = -7490.4042
rho =  0.2     log likelihood =  -7525.883

Iteration 0:   log likelihood = -7480.7796  
Iteration 1:   log likelihood = -7462.5148  
Iteration 2:   log likelihood = -7462.5099  
Iteration 3:   log likelihood = -7462.5099  

Random-effects probit regression                     Number of obs    = 17,634
Group variable: cname                                Number of groups =     20

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    190
                                                                  avg =  881.7
                                                                  max =  1,990

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(11)    =  36.56
Log likelihood = -7462.5099                          Prob > chi2      = 0.0001

----------------------------------------------------------------------------------------
               demosup | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0288271   .0088155     3.27   0.001     .0115491    .0461051
               educlvl |    .047717   .0121431     3.93   0.000      .023917     .071517
                       |
c.lnfdi_jobs#c.educlvl |  -.0054236   .0017839    -3.04   0.002      -.00892   -.0019273
                       |
                   age |   .0030783   .0009252     3.33   0.001     .0012649    .0048917
                female |   .0021425   .0240992     0.09   0.929     -.045091    .0493761
                unempl |   .0160907   .0264496     0.61   0.543    -.0357496     .067931
                 urban |    .027594   .0294955     0.94   0.350     -.030216    .0854041
              interest |   .0018654   .0119715     0.16   0.876    -.0215984    .0253292
                  news |   .0091661   .0089264     1.03   0.304    -.0083293    .0266615
                       |
                 round |
                    5  |  -.0486699   .0381598    -1.28   0.202    -.1234617    .0261218
                    6  |   -.061755    .039237    -1.57   0.116    -.1386581    .0151482
                       |
                 _cons |   .6224175   .1076474     5.78   0.000     .4114326    .8334025
-----------------------+----------------------------------------------------------------
              /lnsig2u |  -2.111447   .3280024                     -2.754319   -1.468574
-----------------------+----------------------------------------------------------------
               sigma_u |   .3479407   .0570627                      .2522941    .4798476
                   rho |   .1079892   .0315957                      .0598432    .1871595
----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 680.04                 Prob >= chibar2 = 0.000

. margins, post ///
>         at(educlvl=1.99 lnfdi_jobs=2.44) ///
>         at(educlvl=1.99 lnfdi_jobs=8.42) ///
>         at(educlvl=6.31 lnfdi_jobs=2.44) ///
>         at(educlvl=6.31 lnfdi_jobs=8.42) atmeans

Adjusted predictions                                    Number of obs = 17,634
Model VCE: OIM

Expression: Pr(demosup=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7940409   .0225059    35.28   0.000     .7499301    .8381516
          2  |    .821834   .0207874    39.54   0.000     .7810915    .8625765
          3  |   .8317786   .0200533    41.48   0.000     .7924749    .8710823
          4  |   .8240062   .0203365    40.52   0.000     .7841475     .863865
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 17,634
Model VCE: OIM

Expression: Pr(demosup=1), predict(pr)
1._at: lnfdi_jobs =     2.44
       educlvl    =     1.99
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)
2._at: lnfdi_jobs =     8.42
       educlvl    =     1.99
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)
3._at: lnfdi_jobs =     2.44
       educlvl    =     6.31
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)
4._at: lnfdi_jobs =     8.42
       educlvl    =     6.31
       age        = 35.45333 (mean)
       female     = .4780538 (mean)
       unempl     = .3039016 (mean)
       urban      = .7202563 (mean)
       interest   = 1.721901 (mean)
       news       = 1.613531 (mean)
       4.round    = .1743223 (mean)
       5.round    = .4169786 (mean)
       6.round    = .4086991 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .7940409  _b[1bn._at]
          2  |    .821834  _b[2._at]
          3  |   .8317786  _b[3._at]
          4  |   .8240062  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0277932   .0095276     2.92   0.004     .0091194    .0464669
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    8.51
         Prob > chi2 =    0.0035

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0077723    .008594    -0.90   0.366    -.0246163    .0090716
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    0.82
         Prob > chi2 =    0.3658

. 
. *** Testing the causal mechanism
. * Mechanism: Economic grievances --> Low legitimacy
. xtreg trustelec problem condlive condecon c.lnfdi_jobs c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,115
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0497                                         min =        196
     Between = 0.3398                                         avg =      953.4
     Overall = 0.0605                                         max =      2,229

                                                Wald chi2(13)     =     953.07
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
   trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
     problem |    .016456   .0071921     2.29   0.022     .0023597    .0305523
    condlive |   .0207883   .0036402     5.71   0.000     .0136536     .027923
    condecon |   .0646163   .0035181    18.37   0.000     .0577209    .0715117
  lnfdi_jobs |   .0005085   .0013882     0.37   0.714    -.0022123    .0032294
     educlvl |   .0103792   .0019914     5.21   0.000     .0064761    .0142823
         age |  -.0020267   .0002773    -7.31   0.000    -.0025702   -.0014831
      female |  -.0022862   .0072284    -0.32   0.752    -.0164536    .0118811
      unempl |   -.001567   .0080365    -0.19   0.845    -.0173181    .0141842
       urban |   .0491911   .0089651     5.49   0.000     .0316197    .0667624
    interest |   -.025738   .0034804    -7.40   0.000    -.0325594   -.0189167
        news |   .0092358   .0027386     3.37   0.001     .0038682    .0146034
             |
       round |
          5  |   .0146238   .0109007     1.34   0.180    -.0067413    .0359889
          6  |  -.0303841   .0109212    -2.78   0.005    -.0517892    -.008979
             |
       _cons |   .2916269    .031062     9.39   0.000     .2307465    .3525074
-------------+----------------------------------------------------------------
     sigma_u |    .093707
     sigma_e |  .46888804
         rho |  .03840586   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. xtreg trustelec problem lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,331
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0171                                         min =        196
     Between = 0.2070                                         avg =      964.8
     Overall = 0.0212                                         max =      2,248

                                                Wald chi2(11)     =     323.53
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
   trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
     problem |   .0220282   .0072705     3.03   0.002     .0077782    .0362782
  lnfdi_jobs |   .0021986   .0014022     1.57   0.117    -.0005497    .0049468
     educlvl |   .0089464   .0020048     4.46   0.000     .0050171    .0128757
         age |  -.0016639   .0002794    -5.95   0.000    -.0022116   -.0011162
      female |  -.0078452   .0073088    -1.07   0.283    -.0221703    .0064799
      unempl |  -.0081835   .0081073    -1.01   0.313    -.0240734    .0077064
       urban |   .0533532   .0090587     5.89   0.000     .0355986    .0711079
    interest |  -.0300003   .0035157    -8.53   0.000     -.036891   -.0231096
        news |   .0068991   .0027603     2.50   0.012      .001489    .0123093
             |
       round |
          5  |   .0335683    .010985     3.06   0.002      .012038    .0550986
          6  |  -.0348351   .0110295    -3.16   0.002    -.0564525   -.0132176
             |
       _cons |   .5722624   .0277657    20.61   0.000     .5178426    .6266822
-------------+----------------------------------------------------------------
     sigma_u |  .08748074
     sigma_e |  .47698839
         rho |  .03254179   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech1

. xtreg trustelec condlive lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,514
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0327                                         min =        197
     Between = 0.2613                                         avg =      974.4
     Overall = 0.0363                                         max =      2,250

                                                Wald chi2(11)     =     630.72
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
   trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
    condlive |    .054789   .0031526    17.38   0.000     .0486101    .0609679
  lnfdi_jobs |   .0016866   .0013855     1.22   0.223    -.0010289    .0044022
     educlvl |    .012096   .0019853     6.09   0.000     .0082049     .015987
         age |  -.0020427   .0002758    -7.41   0.000    -.0025833   -.0015021
      female |  -.0041169     .00722    -0.57   0.569    -.0182678    .0100341
      unempl |    .000346   .0080091     0.04   0.966    -.0153515    .0160435
       urban |   .0552507   .0089363     6.18   0.000      .037736    .0727655
    interest |  -.0293196   .0034699    -8.45   0.000    -.0361205   -.0225187
        news |   .0110849   .0027319     4.06   0.000     .0057304    .0164393
             |
       round |
          5  |   .0245961   .0108512     2.27   0.023     .0033282     .045864
          6  |  -.0308673    .010885    -2.84   0.005    -.0522016   -.0095331
             |
       _cons |   .3952959   .0283969    13.92   0.000     .3396391    .4509527
-------------+----------------------------------------------------------------
     sigma_u |  .08127874
     sigma_e |  .47328805
         rho |  .02864705   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech2

. xtreg trustelec condecon lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,405
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0484                                         min =        197
     Between = 0.3449                                         avg =      968.7
     Overall = 0.0606                                         max =      2,245

                                                Wald chi2(11)     =     942.99
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
   trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
    condecon |   .0751906   .0030281    24.83   0.000     .0692556    .0811255
  lnfdi_jobs |   .0005964   .0013789     0.43   0.665    -.0021062    .0032989
     educlvl |    .009423   .0019684     4.79   0.000     .0055652    .0132809
         age |  -.0020088   .0002744    -7.32   0.000    -.0025467    -.001471
      female |  -.0035122   .0071796    -0.49   0.625    -.0175839    .0105594
      unempl |  -.0017549   .0079595    -0.22   0.825    -.0173552    .0138454
       urban |   .0486487   .0088921     5.47   0.000     .0312205    .0660769
    interest |  -.0266707   .0034555    -7.72   0.000    -.0334434    -.019898
        news |   .0084428   .0027071     3.12   0.002      .003137    .0137486
             |
       round |
          5  |   .0143546   .0108113     1.33   0.184    -.0068352    .0355444
          6  |  -.0320815   .0108246    -2.96   0.003    -.0532973   -.0108658
             |
       _cons |   .3360878   .0275906    12.18   0.000     .2820111    .3901645
-------------+----------------------------------------------------------------
     sigma_u |  .07863918
     sigma_e |  .46932775
         rho |  .02730867   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech3

. * Mechanism: Economic grievances --> Low regime support
. xtreg demoqua problem condlive condecon c.lnfdi_jobs c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,666
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0723                                         min =        204
     Between = 0.4318                                         avg =      983.3
     Overall = 0.0923                                         max =      2,103

                                                Wald chi2(13)     =    1540.29
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
     problem |   .0247891   .0123135     2.01   0.044     .0006552    .0489231
    condlive |   .0403227   .0062677     6.43   0.000     .0280381    .0526073
    condecon |   .1567221   .0061059    25.67   0.000     .1447548    .1686894
  lnfdi_jobs |  -.0020448   .0024035    -0.85   0.395    -.0067555     .002666
     educlvl |   .0333849    .003353     9.96   0.000     .0268131    .0399567
         age |  -.0011538   .0004752    -2.43   0.015    -.0020851   -.0002224
      female |  -.0596811    .012372    -4.82   0.000    -.0839298   -.0354324
      unempl |  -.0135067   .0136502    -0.99   0.322    -.0402607    .0132473
       urban |   .0658818   .0153815     4.28   0.000     .0357346     .096029
    interest |  -.0401921   .0060515    -6.64   0.000    -.0520528   -.0283313
        news |   .0146041   .0046307     3.15   0.002      .005528    .0236802
             |
       round |
          5  |   .1160243   .0190096     6.10   0.000     .0787662    .1532825
          6  |   .0802851   .0195518     4.11   0.000     .0419643    .1186059
             |
       _cons |   1.822098   .0621737    29.31   0.000      1.70024    1.943956
-------------+----------------------------------------------------------------
     sigma_u |   .2157999
     sigma_e |  .83596399
         rho |  .06247553   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. xtreg demoqua problem lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,887
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0181                                         min =        204
     Between = 0.2134                                         avg =      994.4
     Overall = 0.0263                                         max =      2,119

                                                Wald chi2(11)     =     369.46
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
     problem |   .0362795   .0126024     2.88   0.004     .0115792    .0609798
  lnfdi_jobs |   .0009713   .0024601     0.39   0.693    -.0038505    .0057931
     educlvl |   .0303805   .0034121     8.90   0.000     .0236929     .037068
         age |   -.000367   .0004849    -0.76   0.449    -.0013174    .0005834
      female |  -.0662797   .0126626    -5.23   0.000    -.0910979   -.0414615
      unempl |  -.0316579   .0139349    -2.27   0.023    -.0589698   -.0043459
       urban |   .0713733   .0157363     4.54   0.000     .0405306    .1022159
    interest |   -.049888   .0061898    -8.06   0.000    -.0620198   -.0377562
        news |   .0090983   .0047287     1.92   0.054    -.0001698    .0183663
             |
       round |
          5  |   .1578271   .0193924     8.14   0.000     .1198187    .1958354
          6  |   .0669986   .0199987     3.35   0.001     .0278019    .1061954
             |
       _cons |   2.477182   .0614405    40.32   0.000     2.356761    2.597603
-------------+----------------------------------------------------------------
     sigma_u |  .22895119
     sigma_e |  .86059274
         rho |  .06609849   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech4

. xtreg demoqua condlive lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     20,072
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0409                                         min =        205
     Between = 0.2715                                         avg =    1,003.6
     Overall = 0.0481                                         max =      2,123

                                                Wald chi2(11)     =     858.29
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
    condlive |   .1208956   .0054756    22.08   0.000     .1101637    .1316275
  lnfdi_jobs |  -.0002061   .0024236    -0.09   0.932    -.0049562     .004544
     educlvl |    .036921   .0033721    10.95   0.000     .0303117    .0435303
         age |   -.001058   .0004772    -2.22   0.027    -.0019933   -.0001227
      female |   -.057756   .0124685    -4.63   0.000    -.0821938   -.0333182
      unempl |  -.0134271   .0137247    -0.98   0.328    -.0403272    .0134729
       urban |   .0784191   .0154858     5.06   0.000     .0480675    .1087708
    interest |  -.0472205   .0060893    -7.75   0.000    -.0591553   -.0352857
        news |   .0183185   .0046636     3.93   0.000      .009178    .0274591
             |
       round |
          5  |   .1420417   .0191023     7.44   0.000     .1046019    .1794816
          6  |   .0788665   .0196726     4.01   0.000      .040309    .1174241
             |
       _cons |    2.07594   .0626096    33.16   0.000     1.953228    2.198653
-------------+----------------------------------------------------------------
     sigma_u |  .22312532
     sigma_e |  .85079375
         rho |  .06435194   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech5

. xtreg demoqua condecon lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,966
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0702                                         min =        205
     Between = 0.4416                                         avg =      998.3
     Overall = 0.0913                                         max =      2,120

                                                Wald chi2(11)     =    1514.84
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
    condecon |   .1766368   .0052628    33.56   0.000     .1663219    .1869516
  lnfdi_jobs |  -.0023059   .0023931    -0.96   0.335    -.0069963    .0023846
     educlvl |   .0310296   .0033168     9.36   0.000     .0245289    .0375303
         age |  -.0010465   .0004714    -2.22   0.026    -.0019704   -.0001227
      female |  -.0592167   .0123074    -4.81   0.000    -.0833389   -.0350946
      unempl |  -.0199926   .0135406    -1.48   0.140    -.0465317    .0065465
       urban |   .0641489   .0152948     4.19   0.000     .0341716    .0941261
    interest |  -.0409435   .0060194    -6.80   0.000    -.0527413   -.0291457
        news |   .0125025   .0045894     2.72   0.006     .0035075    .0214975
             |
       round |
          5  |   .1179764   .0189158     6.24   0.000     .0809021    .1550507
          6  |   .0805159   .0194285     4.14   0.000     .0424368    .1185951
             |
       _cons |   1.908303    .059403    32.12   0.000     1.791875     2.02473
-------------+----------------------------------------------------------------
     sigma_u |  .20919447
     sigma_e |  .83786959
         rho |  .05867933   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech6

. * Mechanism: Low legitimacy --> Low regime support
. xtreg demoqua trustelec trustlead trustpart c.lnfdi_jobs c.educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     15,806
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.1213                                         min =        187
     Between = 0.8409                                         avg =      878.1
     Overall = 0.1768                                         max =      2,027

                                                Wald chi2(13)     =    2206.37
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
   trustelec |   .2292985   .0160617    14.28   0.000     .1978181    .2607788
   trustlead |   .2754461   .0162917    16.91   0.000      .243515    .3073773
   trustpart |   .2259044   .0169492    13.33   0.000     .1926845    .2591243
  lnfdi_jobs |    .004609   .0025326     1.82   0.069    -.0003548    .0095728
     educlvl |    .023185    .003681     6.30   0.000     .0159704    .0303997
         age |   .0014821   .0005126     2.89   0.004     .0004775    .0024867
      female |  -.0644028    .013209    -4.88   0.000    -.0902919   -.0385138
      unempl |  -.0264784   .0146892    -1.80   0.071    -.0552686    .0023119
       urban |   .0319772   .0168423     1.90   0.058    -.0010332    .0649875
    interest |  -.0345847   .0064142    -5.39   0.000    -.0471564    -.022013
        news |  -.0029621   .0049576    -0.60   0.550    -.0126788    .0067547
             |
       round |
          5  |   .1195554   .0192796     6.20   0.000      .081768    .1573427
          6  |   .0893662   .0198932     4.49   0.000     .0503763    .1283562
             |
       _cons |   2.052133   .0535858    38.30   0.000     1.947107     2.15716
-------------+----------------------------------------------------------------
     sigma_u |  .16924085
     sigma_e |  .80146457
         rho |    .042687   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. xtreg demoqua trustelec lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     17,401
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0839                                         min =        189
     Between = 0.5502                                         avg =      915.8
     Overall = 0.1133                                         max =      2,066

                                                Wald chi2(11)     =    1603.02
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
   trustelec |   .4649701   .0131721    35.30   0.000     .4391533    .4907869
  lnfdi_jobs |   .0049076   .0025024     1.96   0.050     3.00e-06    .0098123
     educlvl |   .0280285   .0035751     7.84   0.000     .0210214    .0350355
         age |    .000709   .0005017     1.41   0.158    -.0002743    .0016923
      female |  -.0639688   .0129851    -4.93   0.000    -.0894192   -.0385185
      unempl |  -.0256414   .0144424    -1.78   0.076     -.053948    .0026651
       urban |   .0533418   .0162853     3.28   0.001     .0214232    .0852603
    interest |  -.0432974   .0063082    -6.86   0.000    -.0556612   -.0309336
        news |   .0020404   .0048864     0.42   0.676    -.0075368    .0116177
             |
       round |
          5  |   .1303338   .0194561     6.70   0.000     .0922007     .168467
          6  |   .0892644   .0198278     4.50   0.000     .0504025    .1281262
             |
       _cons |   2.187764   .0601833    36.35   0.000     2.069807    2.305721
-------------+----------------------------------------------------------------
     sigma_u |  .21422322
     sigma_e |     .82671
         rho |  .06292199   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech7

. xtreg demoqua trustlead lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     19,266
Group variable: cname                           Number of groups  =         20

R-squared:                                      Obs per group:
     Within  = 0.0908                                         min =        199
     Between = 0.6348                                         avg =      963.3
     Overall = 0.1246                                         max =      2,091

                                                Wald chi2(11)     =    1935.44
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
   trustlead |   .4969175   .0125398    39.63   0.000       .47234     .521495
  lnfdi_jobs |  -.0000714   .0023873    -0.03   0.976    -.0047505    .0046076
     educlvl |   .0249313   .0033195     7.51   0.000     .0184252    .0314375
         age |   .0006044   .0004732     1.28   0.202    -.0003231    .0015319
      female |  -.0673693   .0123649    -5.45   0.000    -.0916041   -.0431345
      unempl |  -.0393043   .0135546    -2.90   0.004    -.0658708   -.0127378
       urban |   .0377419   .0154705     2.44   0.015     .0074203    .0680635
    interest |  -.0376111   .0060683    -6.20   0.000    -.0495047   -.0257174
        news |   .0033222   .0046102     0.72   0.471    -.0057137    .0123581
             |
       round |
          5  |   .1385817   .0187976     7.37   0.000      .101739    .1754244
          6  |   .0807608   .0193824     4.17   0.000      .042772    .1187496
             |
       _cons |   2.285459   .0514379    44.43   0.000     2.184643    2.386275
-------------+----------------------------------------------------------------
     sigma_u |  .17744665
     sigma_e |  .82610602
         rho |  .04410368   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech8

. xtreg demoqua trustpart lnfdi_jobs educlvl $C1 $C2 i.round, i(cname)

Random-effects GLS regression                   Number of obs     =     18,166
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0832                                         min =        197
     Between = 0.7779                                         avg =      956.1
     Overall = 0.1266                                         max =      2,089

                                                Wald chi2(11)     =    1672.28
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

------------------------------------------------------------------------------
     demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
   trustpart |   .4832042   .0132676    36.42   0.000     .4572002    .5092083
  lnfdi_jobs |   .0007522   .0024511     0.31   0.759    -.0040519    .0055564
     educlvl |   .0217938   .0034536     6.31   0.000     .0150249    .0285628
         age |   .0005483   .0004882     1.12   0.261    -.0004086    .0015052
      female |  -.0626085    .012701    -4.93   0.000     -.087502    -.037715
      unempl |  -.0362165   .0139684    -2.59   0.010    -.0635941    -.008839
       urban |   .0444601   .0160328     2.77   0.006     .0130363    .0758839
    interest |  -.0376979   .0062254    -6.06   0.000    -.0498994   -.0254963
        news |   .0018551    .004723     0.39   0.694    -.0074018    .0111121
             |
       round |
          5  |   .1420263   .0188924     7.52   0.000     .1049979    .1790547
          6  |    .093076   .0196711     4.73   0.000     .0545214    .1316306
             |
       _cons |    2.19329   .0491556    44.62   0.000     2.096947    2.289633
-------------+----------------------------------------------------------------
     sigma_u |  .15365466
     sigma_e |   .8245098
         rho |  .03356394   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. est store mech9

. 
. *** TABLE A-10: Testing the Causal Mechanism
. esttab mech1 mech2 mech3 mech4 mech5 mech6 mech7 mech8 mech9, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(problem condlive condecon trustelec trustlead trustpart educlvl lnfdi_jobs age female unempl urban i
> nterest news) ///
>         drop(*round _cons) eqlabels("" "") nogaps ///
>         coeflabel(educlvl "Education level" ///
>                 lnfdi_jobs "FDI jobs created" ///
>                 age "Age in years" female "Female" unempl "Unemployed" ///
>                 urban "Urban resident" interest "Political interest" ///
>                 news "Newspaper consumption")

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                trustelec       trustelec       trustelec         demoqua         demoqua         demoqua         de
> moqua         demoqua         demoqua   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
problem             0.022***                                        0.036***                                        
>                                         
                   (0.01)                                          (0.01)                                           
>                                         
condlive                            0.055***                                        0.121***                        
>                                         
                                   (0.00)                                          (0.01)                           
>                                         
condecon                                            0.075***                                        0.177***        
>                                         
                                                   (0.00)                                          (0.01)           
>                                         
trustelec                                                                                                           
> 0.465***                                
                                                                                                                   (
> 0.01)                                   
trustlead                                                                                                           
>                 0.497***                
                                                                                                                    
>                (0.01)                   
trustpart                                                                                                           
>                                 0.483***
                                                                                                                    
>                                (0.01)   
Education ~l        0.009***        0.012***        0.009***        0.030***        0.037***        0.031***        
> 0.028***        0.025***        0.022***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
FDI jobs c~d        0.002           0.002           0.001           0.001          -0.000          -0.002           
> 0.005**        -0.000           0.001   
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Age in years       -0.002***       -0.002***       -0.002***       -0.000          -0.001**        -0.001**         
> 0.001           0.001           0.001   
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
Female             -0.008          -0.004          -0.004          -0.066***       -0.058***       -0.059***       -
> 0.064***       -0.067***       -0.063***
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Unemployed         -0.008           0.000          -0.002          -0.032**        -0.013          -0.020          -
> 0.026*         -0.039***       -0.036***
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Urban resi~t        0.053***        0.055***        0.049***        0.071***        0.078***        0.064***        
> 0.053***        0.038**         0.044***
                   (0.01)          (0.01)          (0.01)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
Political ~t       -0.030***       -0.029***       -0.027***       -0.050***       -0.047***       -0.041***       -
> 0.043***       -0.038***       -0.038***
                   (0.00)          (0.00)          (0.00)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
Newspaper ~n        0.007**         0.011***        0.008***        0.009*          0.018***        0.013***        
> 0.002           0.003           0.002   
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        18331           18514           18405           19887           20072           19966           
> 17401           19266           18166   
\# of coun~s           19              19              19              20              20              20           
>    19              20              19   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** Sensitivity to country selection
. * Leave out one country test: Self-Reported Economic Problems
. foreach i in 1 4 6 8 9 11 12 16 18 20 21 22 24 29 30 31 32 33 34 36 {
  2.         xtreg problem c.lnfdi_jobs##c.educlvl $C1 $C2 i.round if cname!=`i', i(cname)
  3.         est store problem_`i'
  4. }

Random-effects GLS regression                   Number of obs     =     20,699
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0110                                         min =        218
     Between = 0.0114                                         avg =    1,089.4
     Overall = 0.0079                                         max =      2,334

                                                Wald chi2(11)     =     228.98
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0132927   .0025125     5.29   0.000     .0083684     .018217
               educlvl |   .0224999    .003504     6.42   0.000     .0156322    .0293676
                       |
c.lnfdi_jobs#c.educlvl |  -.0029948   .0005202    -5.76   0.000    -.0040144   -.0019752
                       |
                   age |  -.0014862   .0002642    -5.63   0.000     -.002004   -.0009684
                female |   .0049967    .006997     0.71   0.475    -.0087172    .0187105
                unempl |  -.0269247   .0076709    -3.51   0.000    -.0419594     -.01189
                 urban |   .0402514   .0086233     4.67   0.000     .0233501    .0571526
              interest |  -.0083206   .0033637    -2.47   0.013    -.0149133   -.0017279
                  news |   .0062401   .0026499     2.35   0.019     .0010463    .0114339
                       |
                 round |
                    5  |  -.0480622   .0105172    -4.57   0.000    -.0686756   -.0274488
                    6  |  -.0658278   .0107264    -6.14   0.000    -.0868512   -.0448044
                       |
                 _cons |   .4840237   .0358684    13.49   0.000     .4137229    .5543246
-----------------------+----------------------------------------------------------------
               sigma_u |  .12680742
               sigma_e |  .48403581
                   rho |  .06422523   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     21,112
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0104                                         min =        218
     Between = 0.0312                                         avg =    1,111.2
     Overall = 0.0069                                         max =      2,334

                                                Wald chi2(11)     =     219.71
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0129606   .0025052     5.17   0.000     .0080505    .0178706
               educlvl |   .0232379   .0035606     6.53   0.000     .0162592    .0302166
                       |
c.lnfdi_jobs#c.educlvl |   -.002996   .0005199    -5.76   0.000     -.004015    -.001977
                       |
                   age |  -.0015712   .0002609    -6.02   0.000    -.0020826   -.0010598
                female |   .0010215   .0069424     0.15   0.883    -.0125853    .0146283
                unempl |   -.029135   .0075999    -3.83   0.000    -.0440306   -.0142394
                 urban |   .0386521   .0084685     4.56   0.000     .0220542      .05525
              interest |  -.0082006   .0033565    -2.44   0.015    -.0147792   -.0016219
                  news |   .0041963   .0026017     1.61   0.107    -.0009029    .0092956
                       |
                 round |
                    5  |  -.0456674   .0106871    -4.27   0.000    -.0666136   -.0247211
                    6  |  -.0546619   .0108882    -5.02   0.000    -.0760023   -.0333214
                       |
                 _cons |   .4996725   .0355443    14.06   0.000     .4300068    .5693381
-----------------------+----------------------------------------------------------------
               sigma_u |    .124319
               sigma_e |   .4845394
                   rho |  .06176313   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,964
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0113                                         min =        218
     Between = 0.0382                                         avg =    1,103.4
     Overall = 0.0073                                         max =      2,334

                                                Wald chi2(11)     =     237.38
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0133597    .002484     5.38   0.000     .0084911    .0182283
               educlvl |   .0234424   .0035197     6.66   0.000      .016544    .0303409
                       |
c.lnfdi_jobs#c.educlvl |   -.003115   .0005168    -6.03   0.000    -.0041279   -.0021021
                       |
                   age |  -.0016634   .0002608    -6.38   0.000    -.0021746   -.0011522
                female |   .0014213   .0069538     0.20   0.838    -.0122079    .0150504
                unempl |   -.026192   .0076055    -3.44   0.001    -.0410984   -.0112856
                 urban |    .039963   .0084977     4.70   0.000     .0233077    .0566183
              interest |  -.0067985   .0033662    -2.02   0.043    -.0133962   -.0002008
                  news |   .0053551   .0026107     2.05   0.040     .0002383     .010472
                       |
                 round |
                    5  |  -.0528706   .0104954    -5.04   0.000    -.0734412      -.0323
                    6  |  -.0583338   .0107078    -5.45   0.000    -.0793206   -.0373469
                       |
                 _cons |   .4928363   .0334547    14.73   0.000     .4272663    .5584063
-----------------------+----------------------------------------------------------------
               sigma_u |    .113893
               sigma_e |  .48372709
                   rho |  .05252442   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,549
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0130                                         min =        218
     Between = 0.0351                                         avg =    1,081.5
     Overall = 0.0137                                         max =      2,334

                                                Wald chi2(11)     =     271.40
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .015559   .0026785     5.81   0.000     .0103093    .0208087
               educlvl |   .0266206   .0037142     7.17   0.000      .019341    .0339003
                       |
c.lnfdi_jobs#c.educlvl |   -.003291   .0005594    -5.88   0.000    -.0043875   -.0021946
                       |
                   age |  -.0015298   .0002663    -5.74   0.000    -.0020517   -.0010078
                female |   .0013743   .0070433     0.20   0.845    -.0124304     .015179
                unempl |  -.0327656   .0077333    -4.24   0.000    -.0479227   -.0176085
                 urban |    .046839   .0087668     5.34   0.000     .0296564    .0640217
              interest |  -.0073372   .0033784    -2.17   0.030    -.0139588   -.0007155
                  news |   .0046159   .0026526     1.74   0.082    -.0005832     .009815
                       |
                 round |
                    5  |  -.0537067   .0105859    -5.07   0.000    -.0744547   -.0329588
                    6  |  -.0644982   .0108178    -5.96   0.000    -.0857008   -.0432957
                       |
                 _cons |   .4546769   .0330902    13.74   0.000     .3898212    .5195326
-----------------------+----------------------------------------------------------------
               sigma_u |  .10908466
               sigma_e |  .48668096
                   rho |  .04783554   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,958
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0114                                         min =        218
     Between = 0.0026                                         avg =    1,103.1
     Overall = 0.0084                                         max =      2,334

                                                Wald chi2(11)     =     240.51
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0136804   .0024623     5.56   0.000     .0088543    .0185064
               educlvl |    .024194   .0034704     6.97   0.000      .017392    .0309959
                       |
c.lnfdi_jobs#c.educlvl |  -.0031078   .0005107    -6.09   0.000    -.0041087   -.0021068
                       |
                   age |  -.0015806   .0002622    -6.03   0.000    -.0020944   -.0010667
                female |   .0048134   .0069683     0.69   0.490    -.0088442     .018471
                unempl |  -.0273403    .007647    -3.58   0.000    -.0423281   -.0123525
                 urban |   .0355949   .0084456     4.21   0.000      .019042    .0521479
              interest |  -.0079851   .0033693    -2.37   0.018    -.0145887   -.0013814
                  news |   .0059645   .0026183     2.28   0.023     .0008327    .0110963
                       |
                 round |
                    5  |  -.0513845   .0104829    -4.90   0.000    -.0719307   -.0308384
                    6  |  -.0615573   .0106971    -5.75   0.000    -.0825233   -.0405912
                       |
                 _cons |   .4996134   .0357725    13.97   0.000     .4295006    .5697263
-----------------------+----------------------------------------------------------------
               sigma_u |  .12681048
               sigma_e |  .48468727
                   rho |  .06406665   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     21,341
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0111                                         min =        218
     Between = 0.0411                                         avg =    1,123.2
     Overall = 0.0074                                         max =      2,334

                                                Wald chi2(11)     =     237.98
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0132965   .0024591     5.41   0.000     .0084768    .0181162
               educlvl |   .0239491   .0034707     6.90   0.000     .0171467    .0307515
                       |
c.lnfdi_jobs#c.educlvl |  -.0030577   .0005096    -6.00   0.000    -.0040566   -.0020589
                       |
                   age |  -.0015359   .0002607    -5.89   0.000    -.0020469    -.001025
                female |   .0019872   .0068968     0.29   0.773    -.0115303    .0155047
                unempl |  -.0290236   .0075523    -3.84   0.000    -.0438259   -.0142213
                 urban |   .0367479   .0084362     4.36   0.000     .0202133    .0532825
              interest |  -.0079559   .0033452    -2.38   0.017    -.0145124   -.0013994
                  news |   .0054714   .0025895     2.11   0.035     .0003961    .0105468
                       |
                 round |
                    5  |  -.0504007   .0104841    -4.81   0.000    -.0709492   -.0298523
                    6  |  -.0613573   .0106983    -5.74   0.000    -.0823256    -.040389
                       |
                 _cons |   .4996266   .0323788    15.43   0.000     .4361654    .5630878
-----------------------+----------------------------------------------------------------
               sigma_u |  .10847146
               sigma_e |  .48441021
                   rho |  .04774812   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,867
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0108                                         min =        218
     Between = 0.0233                                         avg =    1,098.3
     Overall = 0.0070                                         max =      2,334

                                                Wald chi2(11)     =     225.66
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0127634   .0025135     5.08   0.000      .007837    .0176898
               educlvl |   .0233242   .0034998     6.66   0.000     .0164646    .0301837
                       |
c.lnfdi_jobs#c.educlvl |  -.0030373   .0005176    -5.87   0.000    -.0040519   -.0020228
                       |
                   age |  -.0015276   .0002615    -5.84   0.000    -.0020402    -.001015
                female |    .003944   .0069638     0.57   0.571    -.0097048    .0175928
                unempl |  -.0269897   .0076398    -3.53   0.000    -.0419635    -.012016
                 urban |    .034194   .0084959     4.02   0.000     .0175422    .0508457
              interest |  -.0082456   .0033629    -2.45   0.014    -.0148368   -.0016544
                  news |   .0063747   .0026293     2.42   0.015     .0012213     .011528
                       |
                 round |
                    5  |    -.04957   .0105013    -4.72   0.000    -.0701522   -.0289879
                    6  |  -.0613541   .0107212    -5.72   0.000    -.0823672   -.0403409
                       |
                 _cons |   .4944668   .0355224    13.92   0.000     .4248441    .5640895
-----------------------+----------------------------------------------------------------
               sigma_u |  .12505334
               sigma_e |  .48376171
                   rho |   .0626376   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,710
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0104                                         min =        218
     Between = 0.0359                                         avg =    1,090.0
     Overall = 0.0065                                         max =      2,334

                                                Wald chi2(11)     =     216.80
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0129081   .0024961     5.17   0.000     .0080158    .0178004
               educlvl |   .0228783   .0035186     6.50   0.000      .015982    .0297746
                       |
c.lnfdi_jobs#c.educlvl |  -.0029472   .0005153    -5.72   0.000    -.0039572   -.0019372
                       |
                   age |    -.00158   .0002648    -5.97   0.000     -.002099   -.0010611
                female |  -.0003616    .007004    -0.05   0.959    -.0140893    .0133661
                unempl |  -.0262475   .0076573    -3.43   0.001    -.0412556   -.0112394
                 urban |   .0366815   .0087201     4.21   0.000     .0195904    .0537727
              interest |  -.0073026   .0033775    -2.16   0.031    -.0139224   -.0006829
                  news |    .002681   .0026378     1.02   0.309     -.002489     .007851
                       |
                 round |
                    5  |  -.0598987   .0109088    -5.49   0.000    -.0812795   -.0385179
                    6  |    -.06824   .0111251    -6.13   0.000    -.0900448   -.0464353
                       |
                 _cons |   .5052975    .033576    15.05   0.000     .4394897    .5711054
-----------------------+----------------------------------------------------------------
               sigma_u |  .11300358
               sigma_e |  .48381394
                   rho |  .05173194   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     21,055
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0113                                         min =        218
     Between = 0.0448                                         avg =    1,108.2
     Overall = 0.0070                                         max =      2,334

                                                Wald chi2(11)     =     238.38
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0125877   .0025138     5.01   0.000     .0076607    .0175147
               educlvl |   .0238639   .0035819     6.66   0.000     .0168435    .0308844
                       |
c.lnfdi_jobs#c.educlvl |    -.00299   .0005222    -5.73   0.000    -.0040135   -.0019665
                       |
                   age |  -.0015145   .0002621    -5.78   0.000    -.0020282   -.0010007
                female |   .0001859   .0069526     0.03   0.979     -.013441    .0138127
                unempl |  -.0312389   .0076247    -4.10   0.000    -.0461831   -.0162948
                 urban |   .0369834   .0085187     4.34   0.000     .0202872    .0536797
              interest |  -.0085609   .0033652    -2.54   0.011    -.0151565   -.0019653
                  news |   .0051561   .0026075     1.98   0.048     .0000455    .0102667
                       |
                 round |
                    5  |  -.0524686   .0106627    -4.92   0.000     -.073367   -.0315701
                    6  |   -.065882   .0108462    -6.07   0.000    -.0871402   -.0446237
                       |
                 _cons |   .5096616   .0352851    14.44   0.000     .4405041     .578819
-----------------------+----------------------------------------------------------------
               sigma_u |  .12259754
               sigma_e |  .48503197
                   rho |  .06005189   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,387
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0112                                         min =        218
     Between = 0.0310                                         avg =    1,073.0
     Overall = 0.0075                                         max =      2,334

                                                Wald chi2(11)     =     230.03
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0130935   .0025892     5.06   0.000     .0080187    .0181682
               educlvl |   .0247839   .0035444     6.99   0.000     .0178371    .0317308
                       |
c.lnfdi_jobs#c.educlvl |  -.0032589   .0005421    -6.01   0.000    -.0043213   -.0021964
                       |
                   age |  -.0014399   .0002654    -5.43   0.000    -.0019601   -.0009198
                female |   .0043691   .0070588     0.62   0.536     -.009466    .0182041
                unempl |   -.025225   .0077566    -3.25   0.001    -.0404277   -.0100224
                 urban |     .03884   .0086429     4.49   0.000     .0219002    .0557798
              interest |  -.0066769   .0033845    -1.97   0.049    -.0133104   -.0000434
                  news |   .0070859   .0026574     2.67   0.008     .0018774    .0122944
                       |
                 round |
                    5  |  -.0548745    .010542    -5.21   0.000    -.0755363   -.0342126
                    6  |  -.0547626   .0107636    -5.09   0.000    -.0758589   -.0336664
                       |
                 _cons |   .4725593   .0340851    13.86   0.000     .4057538    .5393647
-----------------------+----------------------------------------------------------------
               sigma_u |  .11673214
               sigma_e |   .4846308
                   rho |  .05483603   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,097
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0115                                         min =        218
     Between = 0.0367                                         avg =    1,057.7
     Overall = 0.0071                                         max =      2,334

                                                Wald chi2(11)     =     231.55
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .011429   .0025617     4.46   0.000     .0064083    .0164498
               educlvl |   .0230622   .0035394     6.52   0.000     .0161252    .0299992
                       |
c.lnfdi_jobs#c.educlvl |   -.002927   .0005222    -5.60   0.000    -.0039506   -.0019034
                       |
                   age |  -.0015222   .0002658    -5.73   0.000    -.0020432   -.0010012
                female |   .0062696   .0071018     0.88   0.377    -.0076498     .020189
                unempl |  -.0327794   .0077368    -4.24   0.000    -.0479433   -.0176155
                 urban |   .0410648   .0086794     4.73   0.000     .0240535    .0580762
              interest |  -.0080347   .0034522    -2.33   0.020    -.0148008   -.0012685
                  news |   .0052479   .0026731     1.96   0.050     8.75e-06     .010487
                       |
                 round |
                    5  |  -.0472206   .0109672    -4.31   0.000     -.068716   -.0257253
                    6  |  -.0637608   .0111749    -5.71   0.000    -.0856631   -.0418584
                       |
                 _cons |   .4993524   .0360083    13.87   0.000     .4287774    .5699275
-----------------------+----------------------------------------------------------------
               sigma_u |  .12568052
               sigma_e |   .4831187
                   rho |   .0633854   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,263
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0114                                         min =        218
     Between = 0.0375                                         avg =    1,066.5
     Overall = 0.0068                                         max =      2,334

                                                Wald chi2(11)     =     232.00
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0145381   .0025136     5.78   0.000     .0096116    .0194647
               educlvl |   .0243893   .0035275     6.91   0.000     .0174754    .0313031
                       |
c.lnfdi_jobs#c.educlvl |  -.0030944   .0005199    -5.95   0.000    -.0041134   -.0020755
                       |
                   age |  -.0015085   .0002665    -5.66   0.000    -.0020308   -.0009863
                female |   .0022054   .0071019     0.31   0.756    -.0117142    .0161249
                unempl |  -.0301228   .0076632    -3.93   0.000    -.0451425   -.0151032
                 urban |   .0308693    .008641     3.57   0.000     .0139332    .0478054
              interest |  -.0081086   .0034305    -2.36   0.018    -.0148322   -.0013849
                  news |   .0055211    .002645     2.09   0.037     .0003369    .0107053
                       |
                 round |
                    5  |  -.0568498   .0111092    -5.12   0.000    -.0786234   -.0350762
                    6  |  -.0723654   .0113474    -6.38   0.000    -.0946059    -.050125
                       |
                 _cons |   .4902828   .0335105    14.63   0.000     .4246035    .5559621
-----------------------+----------------------------------------------------------------
               sigma_u |  .11258052
               sigma_e |    .484191
                   rho |  .05128931   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,613
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0117                                         min =        218
     Between = 0.0095                                         avg =    1,032.3
     Overall = 0.0086                                         max =      2,334

                                                Wald chi2(11)     =     230.35
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0148804   .0025442     5.85   0.000     .0098938     .019867
               educlvl |   .0229665   .0036321     6.32   0.000     .0158477    .0300853
                       |
c.lnfdi_jobs#c.educlvl |  -.0031571   .0005356    -5.89   0.000    -.0042068   -.0021073
                       |
                   age |  -.0015924   .0002679    -5.94   0.000    -.0021175   -.0010674
                female |  -.0006061   .0071675    -0.08   0.933    -.0146543     .013442
                unempl |  -.0312535   .0078262    -3.99   0.000    -.0465926   -.0159144
                 urban |   .0354695   .0088199     4.02   0.000     .0181828    .0527562
              interest |  -.0079431   .0034801    -2.28   0.022    -.0147639   -.0011223
                  news |   .0061452   .0027009     2.28   0.023     .0008515    .0114389
                       |
                 round |
                    5  |  -.0533134   .0115256    -4.63   0.000    -.0759031   -.0307237
                    6  |  -.0695101   .0116757    -5.95   0.000     -.092394   -.0466261
                       |
                 _cons |   .4995513   .0343845    14.53   0.000     .4321589    .5669438
-----------------------+----------------------------------------------------------------
               sigma_u |   .1165112
               sigma_e |  .48265491
                   rho |  .05506358   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     21,426
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0112                                         min =        303
     Between = 0.0254                                         avg =    1,127.7
     Overall = 0.0076                                         max =      2,334

                                                Wald chi2(11)     =     239.52
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0132944   .0024599     5.40   0.000     .0084731    .0181156
               educlvl |   .0238908   .0034487     6.93   0.000     .0171315    .0306502
                       |
c.lnfdi_jobs#c.educlvl |  -.0030356   .0005105    -5.95   0.000    -.0040361   -.0020351
                       |
                   age |  -.0015447   .0002588    -5.97   0.000    -.0020519   -.0010375
                female |   .0030481   .0068852     0.44   0.658    -.0104466    .0165428
                unempl |  -.0299802   .0075269    -3.98   0.000    -.0447327   -.0152277
                 urban |   .0366043    .008422     4.35   0.000     .0200976    .0531111
              interest |  -.0079314   .0033176    -2.39   0.017    -.0144339    -.001429
                  news |   .0048457   .0025889     1.87   0.061    -.0002285    .0099199
                       |
                 round |
                    5  |  -.0510612   .0104712    -4.88   0.000    -.0715843    -.030538
                    6  |  -.0608136    .010681    -5.69   0.000     -.081748   -.0398792
                       |
                 _cons |   .4882928   .0294523    16.58   0.000     .4305672    .5460183
-----------------------+----------------------------------------------------------------
               sigma_u |  .09175585
               sigma_e |  .48410331
                   rho |  .03467874   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,813
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0106                                         min =        218
     Between = 0.0138                                         avg =    1,095.4
     Overall = 0.0074                                         max =      2,334

                                                Wald chi2(11)     =     222.87
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0135388   .0024665     5.49   0.000     .0087046     .018373
               educlvl |   .0246431   .0034945     7.05   0.000     .0177941    .0314922
                       |
c.lnfdi_jobs#c.educlvl |  -.0030725    .000512    -6.00   0.000    -.0040759    -.002069
                       |
                   age |  -.0014873   .0002632    -5.65   0.000    -.0020032   -.0009715
                female |   .0020104   .0069807     0.29   0.773    -.0116714    .0156923
                unempl |  -.0269003    .007612    -3.53   0.000    -.0418197    -.011981
                 urban |   .0360098   .0086654     4.16   0.000     .0190259    .0529937
              interest |  -.0066386   .0033772    -1.97   0.049    -.0132579   -.0000194
                  news |   .0039812   .0026194     1.52   0.129    -.0011527    .0091151
                       |
                 round |
                    5  |  -.0499162   .0104948    -4.76   0.000    -.0704856   -.0293468
                    6  |  -.0633676   .0107184    -5.91   0.000    -.0843753   -.0423599
                       |
                 _cons |   .4849851   .0352796    13.75   0.000     .4158383    .5541318
-----------------------+----------------------------------------------------------------
               sigma_u |  .12389291
               sigma_e |    .483768
                   rho |  .06155023   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,140
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0112                                         min =        218
     Between = 0.0250                                         avg =    1,060.0
     Overall = 0.0066                                         max =      2,334

                                                Wald chi2(11)     =     225.60
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0134946   .0025374     5.32   0.000     .0085215    .0184677
               educlvl |   .0229522   .0035132     6.53   0.000     .0160665    .0298379
                       |
c.lnfdi_jobs#c.educlvl |  -.0030648   .0005182    -5.91   0.000    -.0040805   -.0020491
                       |
                   age |  -.0017072   .0002692    -6.34   0.000    -.0022347   -.0011797
                female |    .005998    .007116     0.84   0.399    -.0079491    .0199451
                unempl |  -.0298494   .0077603    -3.85   0.000    -.0450594   -.0146394
                 urban |   .0278161   .0087796     3.17   0.002     .0106083    .0450239
              interest |  -.0090673   .0034335    -2.64   0.008    -.0157968   -.0023379
                  news |   .0072648     .00269     2.70   0.007     .0019925    .0125371
                       |
                 round |
                    5  |  -.0533573   .0109196    -4.89   0.000    -.0747592   -.0319553
                    6  |  -.0601011   .0111101    -5.41   0.000    -.0818766   -.0383256
                       |
                 _cons |   .5143208   .0332474    15.47   0.000     .4491571    .5794844
-----------------------+----------------------------------------------------------------
               sigma_u |  .11078863
               sigma_e |  .48537351
                   rho |  .04952007   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,837
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0104                                         min =        218
     Between = 0.0297                                         avg =    1,096.7
     Overall = 0.0064                                         max =      2,334

                                                Wald chi2(11)     =     217.01
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0122312   .0024878     4.92   0.000     .0073552    .0171072
               educlvl |    .024054   .0035365     6.80   0.000     .0171226    .0309855
                       |
c.lnfdi_jobs#c.educlvl |  -.0030063   .0005163    -5.82   0.000    -.0040183   -.0019943
                       |
                   age |  -.0015546   .0002624    -5.92   0.000    -.0020689   -.0010403
                female |   .0000893   .0069709     0.01   0.990    -.0135733     .013752
                unempl |  -.0262587    .007662    -3.43   0.001     -.041276   -.0112414
                 urban |   .0330064    .008547     3.86   0.000     .0162546    .0497582
              interest |  -.0066176   .0033801    -1.96   0.050    -.0132426    7.30e-06
                  news |   .0047765   .0026098     1.83   0.067    -.0003386    .0098916
                       |
                 round |
                    5  |  -.0515311   .0104994    -4.91   0.000    -.0721095   -.0309527
                    6  |  -.0576302   .0107181    -5.38   0.000    -.0786373   -.0366231
                       |
                 _cons |    .494707   .0357194    13.85   0.000     .4246982    .5647157
-----------------------+----------------------------------------------------------------
               sigma_u |  .12588811
               sigma_e |  .48385945
                   rho |  .06339944   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     20,215
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0116                                         min =        218
     Between = 0.0064                                         avg =    1,063.9
     Overall = 0.0090                                         max =      2,334

                                                Wald chi2(11)     =     235.81
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0133217   .0025585     5.21   0.000     .0083071    .0183363
               educlvl |   .0244299   .0035025     6.97   0.000     .0175651    .0312946
                       |
c.lnfdi_jobs#c.educlvl |  -.0029981   .0005325    -5.63   0.000    -.0040419   -.0019544
                       |
                   age |  -.0015951   .0002687    -5.94   0.000    -.0021219   -.0010684
                female |   .0042431   .0070702     0.60   0.548    -.0096142    .0181004
                unempl |   -.028455   .0077585    -3.67   0.000    -.0436614   -.0132485
                 urban |      .0377   .0086681     4.35   0.000     .0207108    .0546893
              interest |   -.006505   .0034037    -1.91   0.056    -.0131762    .0001661
                  news |   .0043142   .0026671     1.62   0.106    -.0009132    .0095416
                       |
                 round |
                    5  |  -.0531688    .010522    -5.05   0.000    -.0737915   -.0325461
                    6  |  -.0602666   .0107515    -5.61   0.000    -.0813392    -.039194
                       |
                 _cons |   .4819955   .0359558    13.41   0.000     .4115235    .5524675
-----------------------+----------------------------------------------------------------
               sigma_u |  .12675684
               sigma_e |   .4836711
                   rho |  .06426789   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,310
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0086                                         min =        218
     Between = 0.0023                                         avg =    1,016.3
     Overall = 0.0064                                         max =      2,031

                                                Wald chi2(11)     =     165.97
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0149624   .0025604     5.84   0.000     .0099441    .0199807
               educlvl |   .0248102   .0036793     6.74   0.000     .0175989    .0320216
                       |
c.lnfdi_jobs#c.educlvl |  -.0033518     .00053    -6.32   0.000    -.0043906    -.002313
                       |
                   age |  -.0015076   .0002707    -5.57   0.000    -.0020382   -.0009769
                female |   .0042568     .00725     0.59   0.557    -.0099529    .0184665
                unempl |  -.0292127   .0078547    -3.72   0.000    -.0446076   -.0138177
                 urban |   .0312901   .0090715     3.45   0.001     .0135104    .0490699
              interest |  -.0040248   .0035098    -1.15   0.252    -.0109039    .0028544
                  news |   .0046619   .0027027     1.72   0.085    -.0006354    .0099592
                       |
                 round |
                    5  |  -.0064078   .0117575    -0.55   0.586     -.029452    .0166363
                    6  |  -.0107677   .0119601    -0.90   0.368     -.034209    .0126737
                       |
                 _cons |   .4453554   .0344363    12.93   0.000     .3778614    .5128494
-----------------------+----------------------------------------------------------------
               sigma_u |  .11407888
               sigma_e |  .48399529
                   rho |  .05263166   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,880
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0124                                         min =        218
     Between = 0.0682                                         avg =    1,046.3
     Overall = 0.0067                                         max =      2,334

                                                Wald chi2(11)     =     247.25
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0132067   .0024983     5.29   0.000     .0083102    .0181032
               educlvl |   .0233436   .0035243     6.62   0.000     .0164361    .0302511
                       |
c.lnfdi_jobs#c.educlvl |  -.0031015   .0005176    -5.99   0.000     -.004116   -.0020871
                       |
                   age |  -.0015717   .0002687    -5.85   0.000    -.0020983   -.0010452
                female |    .008892   .0071378     1.25   0.213    -.0050978    .0228817
                unempl |  -.0304342   .0077824    -3.91   0.000    -.0456873    -.015181
                 urban |   .0420191   .0086939     4.83   0.000     .0249794    .0590588
              interest |  -.0087783   .0034349    -2.56   0.011    -.0155107    -.002046
                  news |   .0032312    .002673     1.21   0.227    -.0020077    .0084701
                       |
                 round |
                    5  |  -.0713253   .0111079    -6.42   0.000    -.0930964   -.0495543
                    6  |  -.0869254   .0114058    -7.62   0.000    -.1092803   -.0645705
                       |
                 _cons |   .5106996   .0351681    14.52   0.000     .4417714    .5796278
-----------------------+----------------------------------------------------------------
               sigma_u |  .12181029
               sigma_e |  .48340421
                   rho |  .05970507   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. coefplot problem_1 problem_4 problem_6 problem_8 problem_9 ///
>         problem_11 problem_12 problem_16 problem_18 problem_20 ///
>         problem_21 problem_22 problem_24 problem_29 problem_30 ///
>         problem_31 problem_32 problem_33 problem_34 problem_36, ///
>         keep(lnfdi_jobs c.lnfdi_jobs#c.educlvl) xline(0) ///
>         title("Self-Reported Economic Problems") ///
>         rename(lnfdi_jobs = "FDI exposure" c.lnfdi_jobs#c.educlvl = "FDI * education") ///
>         legend(order(- "Country excluded:" " " 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40) ///
>                 label(2 "Algeria") label(4 "Burkina Faso") label(6 "Cameroon") label(8 "Egypt") label(10 "Gabon") 
> ///
>                 label(12 "Guinea") label(14 "Ivory Coast") label(16 "Madagascar") label(18 "Mali") label(20 "Moroc
> co") ///
>                 label(22 "Mozambique") label(24 "Namibia") label(26 "Nigeria") label(28 "Sudan") label(30 "Swazila
> nd") ///
>                 label(32 "Tanzania") label(34 "Togo") label(36 "Tunisia") label(38 "Uganda") label(40 "Zimbabwe"))
>  ///
>         name(afb_jackknife1, replace)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)

. * Leave out one country test: Distrust in Electoral Commission
. foreach i in 1 4 6 8 9 11 12 16 18 20 21 22 24 29 30 31 32 33 34 36 {
  2.         xtreg trustelec c.lnfdi_jobs##c.educlvl $C1 $C2 i.round if cname!=`i', i(cname)
  3.         est store trustelec_`i'
  4. }

Random-effects GLS regression                   Number of obs     =     17,743
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0176                                         min =        197
     Between = 0.1974                                         avg =      985.7
     Overall = 0.0221                                         max =      2,258

                                                Wald chi2(11)     =     323.04
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .009211   .0028127     3.27   0.001     .0036982    .0147238
               educlvl |   .0193027   .0037306     5.17   0.000     .0119909    .0266145
                       |
c.lnfdi_jobs#c.educlvl |  -.0017415   .0005853    -2.98   0.003    -.0028886   -.0005943
                       |
                   age |  -.0016818   .0002842    -5.92   0.000    -.0022388   -.0011248
                female |  -.0061837    .007433    -0.83   0.405    -.0207521    .0083846
                unempl |  -.0047553   .0082586    -0.58   0.565    -.0209418    .0114312
                 urban |   .0504939   .0092627     5.45   0.000     .0323394    .0686483
              interest |  -.0322879   .0035559    -9.08   0.000    -.0392574   -.0253184
                  news |   .0064758   .0028264     2.29   0.022     .0009361    .0120155
                       |
                 round |
                    5  |   .0331465   .0109766     3.02   0.003     .0116329    .0546602
                    6  |  -.0371836   .0110014    -3.38   0.001    -.0587459   -.0156213
                       |
                 _cons |   .5435273   .0285789    19.02   0.000     .4875137    .5995409
-----------------------+----------------------------------------------------------------
               sigma_u |  .07607904
               sigma_e |  .47707106
                   rho |  .02480033   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,087
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0175                                         min =        197
     Between = 0.1951                                         avg =    1,004.8
     Overall = 0.0218                                         max =      2,258

                                                Wald chi2(11)     =     327.16
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0095798   .0027837     3.44   0.001     .0041238    .0150357
               educlvl |   .0187959   .0037768     4.98   0.000     .0113934    .0261984
                       |
c.lnfdi_jobs#c.educlvl |  -.0017638   .0005793    -3.04   0.002    -.0028991   -.0006284
                       |
                   age |  -.0017008     .00028    -6.07   0.000    -.0022496   -.0011519
                female |  -.0089547   .0073556    -1.22   0.223    -.0233713     .005462
                unempl |  -.0037973   .0081619    -0.47   0.642    -.0197944    .0121997
                 urban |   .0526707   .0090647     5.81   0.000     .0349042    .0704372
              interest |  -.0305883   .0035396    -8.64   0.000    -.0375259   -.0236507
                  news |   .0071902   .0027685     2.60   0.009      .001764    .0126164
                       |
                 round |
                    5  |   .0352395   .0111064     3.17   0.002     .0134713    .0570077
                    6  |  -.0368153    .011133    -3.31   0.001    -.0586356   -.0149951
                       |
                 _cons |   .5454464   .0292134    18.67   0.000     .4881891    .6027037
-----------------------+----------------------------------------------------------------
               sigma_u |  .07948516
               sigma_e |  .47648937
                   rho |  .02707358   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,950
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0179                                         min =        197
     Between = 0.2095                                         avg =      997.2
     Overall = 0.0222                                         max =      2,258

                                                Wald chi2(11)     =     333.12
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0086794   .0027632     3.14   0.002     .0032636    .0140952
               educlvl |   .0172354    .003736     4.61   0.000      .009913    .0245578
                       |
c.lnfdi_jobs#c.educlvl |  -.0015386   .0005775    -2.66   0.008    -.0026704   -.0004069
                       |
                   age |  -.0018398   .0002803    -6.56   0.000    -.0023891   -.0012904
                female |  -.0090273   .0073839    -1.22   0.221    -.0234996    .0054449
                unempl |  -.0051967   .0081816    -0.64   0.525    -.0212324    .0108391
                 urban |   .0530015   .0091088     5.82   0.000     .0351487    .0708544
              interest |  -.0301468   .0035561    -8.48   0.000    -.0371167    -.023177
                  news |   .0080618   .0027832     2.90   0.004     .0026069    .0135168
                       |
                 round |
                    5  |   .0331269   .0109418     3.03   0.002     .0116814    .0545724
                    6  |  -.0365124   .0109742    -3.33   0.001    -.0580214   -.0150034
                       |
                 _cons |   .5540927   .0279605    19.82   0.000     .4992911    .6088943
-----------------------+----------------------------------------------------------------
               sigma_u |  .07250879
               sigma_e |  .47642466
                   rho |  .02263852   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,554
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0183                                         min =        197
     Between = 0.2290                                         avg =      975.2
     Overall = 0.0230                                         max =      2,258

                                                Wald chi2(11)     =     330.96
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0118336   .0030135     3.93   0.000     .0059272      .01774
               educlvl |   .0220677   .0039618     5.57   0.000     .0143028    .0298326
                       |
c.lnfdi_jobs#c.educlvl |  -.0022372   .0006356    -3.52   0.000    -.0034831   -.0009914
                       |
                   age |  -.0018848   .0002848    -6.62   0.000     -.002443   -.0013267
                female |  -.0100535   .0074356    -1.35   0.176     -.024627      .00452
                unempl |  -.0114374    .008281    -1.38   0.167    -.0276679    .0047932
                 urban |   .0503126   .0093777     5.37   0.000     .0319327    .0686925
              interest |  -.0316445   .0035427    -8.93   0.000    -.0385881   -.0247009
                  news |   .0095127   .0028141     3.38   0.001     .0039971    .0150283
                       |
                 round |
                    5  |   .0189211   .0109733     1.72   0.085    -.0025862    .0404285
                    6  |  -.0256601   .0110268    -2.33   0.020    -.0472722    -.004048
                       |
                 _cons |   .5375217   .0299161    17.97   0.000     .4788872    .5961562
-----------------------+----------------------------------------------------------------
               sigma_u |  .08169404
               sigma_e |  .47594436
                   rho |  .02861922   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,892
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0179                                         min =        197
     Between = 0.1744                                         avg =      994.0
     Overall = 0.0210                                         max =      2,258

                                                Wald chi2(11)     =     328.86
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .009917   .0027546     3.60   0.000     .0045182    .0153159
               educlvl |   .0182533   .0037039     4.93   0.000     .0109938    .0255127
                       |
c.lnfdi_jobs#c.educlvl |  -.0017713   .0005735    -3.09   0.002    -.0028953   -.0006473
                       |
                   age |  -.0017607   .0002833    -6.21   0.000     -.002316   -.0012054
                female |  -.0096905   .0074346    -1.30   0.192     -.024262     .004881
                unempl |  -.0095145   .0082748    -1.15   0.250    -.0257328    .0067039
                 urban |   .0513958   .0090885     5.66   0.000     .0335826     .069209
              interest |  -.0304943   .0035781    -8.52   0.000    -.0375072   -.0234814
                  news |   .0076111   .0028046     2.71   0.007     .0021142     .013108
                       |
                 round |
                    5  |   .0318557   .0109532     2.91   0.004     .0103879    .0533235
                    6  |  -.0374014   .0109997    -3.40   0.001    -.0589604   -.0158424
                       |
                 _cons |   .5391759   .0286144    18.84   0.000     .4830926    .5952592
-----------------------+----------------------------------------------------------------
               sigma_u |  .07685441
               sigma_e |  .47914569
                   rho |  .02508247   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,279
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0177                                         min =        197
     Between = 0.1844                                         avg =    1,015.5
     Overall = 0.0216                                         max =      2,258

                                                Wald chi2(11)     =     333.60
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0094772    .002737     3.46   0.001     .0041128    .0148415
               educlvl |   .0182847    .003687     4.96   0.000     .0110583    .0255111
                       |
c.lnfdi_jobs#c.educlvl |  -.0017246   .0005691    -3.03   0.002      -.00284   -.0006091
                       |
                   age |  -.0016345   .0002802    -5.83   0.000    -.0021838   -.0010853
                female |  -.0092484   .0073151    -1.26   0.206    -.0235858     .005089
                unempl |  -.0107374   .0081189    -1.32   0.186    -.0266501    .0051754
                 urban |   .0526192   .0090461     5.82   0.000     .0348892    .0703493
              interest |  -.0311192   .0035336    -8.81   0.000     -.038045   -.0241933
                  news |   .0073988   .0027577     2.68   0.007     .0019938    .0128039
                       |
                 round |
                    5  |   .0339842   .0109213     3.11   0.002      .012579    .0553895
                    6  |  -.0383294   .0109601    -3.50   0.000    -.0598107    -.016848
                       |
                 _cons |   .5453393   .0288497    18.90   0.000      .488795    .6018836
-----------------------+----------------------------------------------------------------
               sigma_u |  .07947881
               sigma_e |   .4768176
                   rho |  .02703312   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,819
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0179                                         min =        197
     Between = 0.2113                                         avg =      989.9
     Overall = 0.0221                                         max =      2,258

                                                Wald chi2(11)     =     331.17
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0087118   .0028137     3.10   0.002      .003197    .0142266
               educlvl |    .016797    .003727     4.51   0.000     .0094923    .0241017
                       |
c.lnfdi_jobs#c.educlvl |  -.0016121   .0005819    -2.77   0.006    -.0027525   -.0004717
                       |
                   age |   -.001723   .0002813    -6.13   0.000    -.0022743   -.0011717
                female |  -.0133916   .0073986    -1.81   0.070    -.0278926    .0011095
                unempl |  -.0065381   .0082295    -0.79   0.427    -.0226677    .0095914
                 urban |   .0594537   .0091091     6.53   0.000     .0416002    .0773073
              interest |  -.0290556   .0035543    -8.17   0.000    -.0360218   -.0220894
                  news |   .0082634   .0028039     2.95   0.003     .0027679    .0137589
                       |
                 round |
                    5  |   .0346625   .0109429     3.17   0.002     .0132148    .0561102
                    6  |  -.0367669   .0109813    -3.35   0.001    -.0582898    -.015244
                       |
                 _cons |   .5498062   .0274054    20.06   0.000     .4960926    .6035198
-----------------------+----------------------------------------------------------------
               sigma_u |  .06814469
               sigma_e |  .47618975
                   rho |  .02006781   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,955
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0184                                         min =        197
     Between = 0.1933                                         avg =      997.5
     Overall = 0.0225                                         max =      2,258

                                                Wald chi2(11)     =     341.83
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0089957   .0027665     3.25   0.001     .0035735     .014418
               educlvl |   .0179282   .0037228     4.82   0.000     .0106316    .0252249
                       |
c.lnfdi_jobs#c.educlvl |  -.0016536    .000574    -2.88   0.004    -.0027787   -.0005286
                       |
                   age |  -.0018728   .0002826    -6.63   0.000    -.0024267   -.0013189
                female |  -.0106673   .0073821    -1.45   0.148    -.0251361    .0038014
                unempl |   -.007863   .0081705    -0.96   0.336    -.0238769    .0081509
                 urban |   .0571487   .0092524     6.18   0.000     .0390142    .0752831
              interest |  -.0317999   .0035428    -8.98   0.000    -.0387436   -.0248562
                  news |   .0055833   .0027865     2.00   0.045     .0001219    .0110447
                       |
                 round |
                    5  |   .0298233   .0110992     2.69   0.007     .0080693    .0515773
                    6  |    -.03909   .0113219    -3.45   0.001    -.0612806   -.0168994
                       |
                 _cons |   .5612428   .0273535    20.52   0.000     .5076308    .6148547
-----------------------+----------------------------------------------------------------
               sigma_u |  .06639024
               sigma_e |    .476124
                   rho |   .0190724   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,033
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0175                                         min =        197
     Between = 0.2019                                         avg =    1,001.8
     Overall = 0.0218                                         max =      2,258

                                                Wald chi2(11)     =     326.29
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0118158   .0027874     4.24   0.000     .0063525    .0172791
               educlvl |   .0217439   .0037934     5.73   0.000     .0143089    .0291789
                       |
c.lnfdi_jobs#c.educlvl |  -.0022208   .0005811    -3.82   0.000    -.0033598   -.0010817
                       |
                   age |  -.0016854   .0002809    -6.00   0.000    -.0022361   -.0011348
                female |   -.008987   .0073621    -1.22   0.222    -.0234164    .0054424
                unempl |  -.0044378   .0081822    -0.54   0.588    -.0204748    .0115991
                 urban |   .0545458   .0091131     5.99   0.000     .0366845    .0724072
              interest |  -.0300768   .0035448    -8.48   0.000    -.0370245   -.0231291
                  news |   .0072722   .0027722     2.62   0.009     .0018388    .0127056
                       |
                 round |
                    5  |   .0322741    .011065     2.92   0.004     .0105871    .0539611
                    6  |  -.0303446   .0110656    -2.74   0.006    -.0520328   -.0086565
                       |
                 _cons |   .5281192   .0286815    18.41   0.000     .4719046    .5843338
-----------------------+----------------------------------------------------------------
               sigma_u |   .0757444
               sigma_e |  .47642204
                   rho |  .02465337   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,583
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0175                                         min =        197
     Between = 0.1990                                         avg =      978.1
     Overall = 0.0217                                         max =      2,258

                                                Wald chi2(11)     =     336.39
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0097377   .0027323     3.56   0.000     .0043825    .0150929
               educlvl |   .0189582   .0036644     5.17   0.000     .0117762    .0261403
                       |
c.lnfdi_jobs#c.educlvl |  -.0017944   .0005684    -3.16   0.002    -.0029084   -.0006804
                       |
                   age |  -.0017365   .0002768    -6.27   0.000     -.002279    -.001194
                female |  -.0082069   .0072625    -1.13   0.258    -.0224411    .0060272
                unempl |  -.0066345   .0080427    -0.82   0.409    -.0223979    .0091289
                 urban |   .0522606   .0090092     5.80   0.000     .0346029    .0699182
              interest |  -.0307532   .0034905    -8.81   0.000    -.0375944    -.023912
                  news |   .0071711   .0027387     2.62   0.009     .0018033    .0125389
                       |
                 round |
                    5  |   .0319239   .0109021     2.93   0.003     .0105561    .0532916
                    6  |  -.0361977   .0109388    -3.31   0.001    -.0576374    -.014758
                       |
                 _cons |   .5458911   .0279087    19.56   0.000     .4911911     .600591
-----------------------+----------------------------------------------------------------
               sigma_u |  .07553513
               sigma_e |  .47701053
                   rho |  .02446168   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,113
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0157                                         min =        197
     Between = 0.1801                                         avg =      950.7
     Overall = 0.0203                                         max =      2,258

                                                Wald chi2(11)     =     277.23
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0076101   .0028737     2.65   0.008     .0019777    .0132425
               educlvl |   .0175362   .0037745     4.65   0.000     .0101383    .0249341
                       |
c.lnfdi_jobs#c.educlvl |  -.0016502   .0005876    -2.81   0.005    -.0028018   -.0004986
                       |
                   age |  -.0017086   .0002867    -5.96   0.000    -.0022705   -.0011467
                female |  -.0092081    .007562    -1.22   0.223    -.0240293    .0056131
                unempl |  -.0117309   .0083492    -1.41   0.160     -.028095    .0046331
                 urban |   .0499546   .0093336     5.35   0.000     .0316612    .0682481
              interest |  -.0278614   .0036548    -7.62   0.000    -.0350248   -.0206981
                  news |    .006447   .0028557     2.26   0.024     .0008499    .0120441
                       |
                 round |
                    5  |   .0213644   .0114591     1.86   0.062    -.0010949    .0438237
                    6  |  -.0453436   .0114655    -3.95   0.000    -.0678156   -.0228716
                       |
                 _cons |    .573916   .0299812    19.14   0.000      .515154     .632678
-----------------------+----------------------------------------------------------------
               sigma_u |   .0818696
               sigma_e |  .47598677
                   rho |  .02873385   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,236
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0161                                         min =        197
     Between = 0.3532                                         avg =      957.6
     Overall = 0.0240                                         max =      2,258

                                                Wald chi2(11)     =     289.36
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0102971   .0028137     3.66   0.000     .0047824    .0158119
               educlvl |   .0197815   .0037638     5.26   0.000     .0124047    .0271584
                       |
c.lnfdi_jobs#c.educlvl |  -.0018557   .0005844    -3.18   0.001     -.003001   -.0007103
                       |
                   age |  -.0015521   .0002875    -5.40   0.000    -.0021156   -.0009885
                female |  -.0060627   .0075714    -0.80   0.423    -.0209024    .0087769
                unempl |  -.0088797   .0082643    -1.07   0.283    -.0250774     .007318
                 urban |   .0461649   .0092968     4.97   0.000     .0279434    .0643863
              interest |  -.0302135   .0036369    -8.31   0.000    -.0373417   -.0230854
                  news |   .0084137   .0028267     2.98   0.003     .0028735    .0139539
                       |
                 round |
                    5  |   .0273158   .0116025     2.35   0.019     .0045753    .0500562
                    6  |  -.0334719   .0116559    -2.87   0.004    -.0563172   -.0106267
                       |
                 _cons |   .5486939   .0285486    19.22   0.000     .4927395    .6046482
-----------------------+----------------------------------------------------------------
               sigma_u |   .0735782
               sigma_e |  .47736386
                   rho |   .0232061   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     16,581
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0191                                         min =        197
     Between = 0.1605                                         avg =      921.2
     Overall = 0.0203                                         max =      2,258

                                                Wald chi2(11)     =     326.46
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0087818   .0028895     3.04   0.002     .0031185     .014445
               educlvl |   .0181387   .0039329     4.61   0.000     .0104304    .0258471
                       |
c.lnfdi_jobs#c.educlvl |   -.001647   .0006125    -2.69   0.007    -.0028475   -.0004464
                       |
                   age |  -.0017981   .0002926    -6.15   0.000    -.0023716   -.0012246
                female |  -.0082478   .0077341    -1.07   0.286    -.0234064    .0069107
                unempl |   -.005077   .0085631    -0.59   0.553    -.0218603    .0117063
                 urban |   .0598675   .0096235     6.22   0.000     .0410059    .0787292
              interest |  -.0305973   .0037337    -8.19   0.000    -.0379152   -.0232793
                  news |   .0059951   .0029263     2.05   0.040     .0002597    .0117305
                       |
                 round |
                    5  |   .0370216   .0122087     3.03   0.002     .0130929    .0609502
                    6  |  -.0404589   .0121304    -3.34   0.001    -.0642341   -.0166838
                       |
                 _cons |   .5407926   .0288657    18.73   0.000     .4842169    .5973683
-----------------------+----------------------------------------------------------------
               sigma_u |  .07255089
               sigma_e |  .48030093
                   rho |    .022308   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,386
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0176                                         min =        304
     Between = 0.1530                                         avg =    1,021.4
     Overall = 0.0214                                         max =      2,258

                                                Wald chi2(11)     =     332.59
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0099593   .0027457     3.63   0.000     .0045779    .0153407
               educlvl |    .019208   .0036703     5.23   0.000     .0120143    .0264018
                       |
c.lnfdi_jobs#c.educlvl |  -.0018698    .000572    -3.27   0.001     -.002991   -.0007487
                       |
                   age |  -.0017005   .0002782    -6.11   0.000    -.0022458   -.0011553
                female |   -.007549    .007308    -1.03   0.302    -.0218724    .0067743
                unempl |  -.0063557   .0080958    -0.79   0.432    -.0222233    .0095118
                 urban |   .0524669   .0090443     5.80   0.000     .0347405    .0701933
              interest |  -.0309082   .0035047    -8.82   0.000    -.0377774   -.0240391
                  news |   .0071765   .0027603     2.60   0.009     .0017664    .0125866
                       |
                 round |
                    5  |   .0321703   .0109261     2.94   0.003     .0107556     .053585
                    6  |  -.0369969   .0109612    -3.38   0.001    -.0584805   -.0155133
                       |
                 _cons |   .5366091   .0286352    18.74   0.000     .4804851    .5927331
-----------------------+----------------------------------------------------------------
               sigma_u |  .07848977
               sigma_e |  .47742207
                   rho |  .02631714   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,804
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0171                                         min =        197
     Between = 0.2037                                         avg =      989.1
     Overall = 0.0215                                         max =      2,258

                                                Wald chi2(11)     =     313.65
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0098342   .0027452     3.58   0.000     .0044537    .0152147
               educlvl |   .0186772   .0037142     5.03   0.000     .0113976    .0259569
                       |
c.lnfdi_jobs#c.educlvl |  -.0018638   .0005719    -3.26   0.001    -.0029847    -.000743
                       |
                   age |  -.0016542   .0002828    -5.85   0.000    -.0022085      -.0011
                female |  -.0064698   .0074112    -0.87   0.383    -.0209955    .0080559
                unempl |  -.0093246   .0081827    -1.14   0.254    -.0253625    .0067132
                 urban |   .0521911    .009321     5.60   0.000     .0339222      .07046
              interest |  -.0311188   .0035669    -8.72   0.000    -.0381098   -.0241277
                  news |   .0067796   .0027915     2.43   0.015     .0013083    .0122509
                       |
                 round |
                    5  |   .0335149   .0109438     3.06   0.002     .0120654    .0549644
                    6  |  -.0366702   .0109865    -3.34   0.001    -.0582032   -.0151371
                       |
                 _cons |   .5516933   .0289738    19.04   0.000     .4949056    .6084809
-----------------------+----------------------------------------------------------------
               sigma_u |  .07920436
               sigma_e |  .47631392
                   rho |  .02690704   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,105
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0180                                         min =        197
     Between = 0.1826                                         avg =      950.3
     Overall = 0.0214                                         max =      2,258

                                                Wald chi2(11)     =     316.94
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .009721   .0028345     3.43   0.001     .0041655    .0152766
               educlvl |   .0183621   .0037327     4.92   0.000     .0110462     .025678
                       |
c.lnfdi_jobs#c.educlvl |  -.0017908   .0005806    -3.08   0.002    -.0029287   -.0006528
                       |
                   age |  -.0017681   .0002901    -6.09   0.000    -.0023368   -.0011995
                female |  -.0053174   .0075595    -0.70   0.482    -.0201337     .009499
                unempl |  -.0034815   .0083553    -0.42   0.677    -.0198577    .0128946
                 urban |   .0572016   .0094368     6.06   0.000     .0387057    .0756975
              interest |  -.0311564   .0036265    -8.59   0.000    -.0382643   -.0240485
                  news |   .0062225   .0028717     2.17   0.030     .0005941     .011851
                       |
                 round |
                    5  |   .0212161   .0113848     1.86   0.062    -.0010978    .0435299
                    6  |  -.0481639   .0113692    -4.24   0.000    -.0704472   -.0258806
                       |
                 _cons |   .5649942   .0283638    19.92   0.000      .509402    .6205863
-----------------------+----------------------------------------------------------------
               sigma_u |  .07207369
               sigma_e |  .47653399
                   rho |  .02236367   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,828
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0180                                         min =        197
     Between = 0.1983                                         avg =      990.4
     Overall = 0.0223                                         max =      2,258

                                                Wald chi2(11)     =     330.55
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0092789   .0027756     3.34   0.001     .0038389    .0147189
               educlvl |   .0206465   .0037672     5.48   0.000     .0132629      .02803
                       |
c.lnfdi_jobs#c.educlvl |  -.0019018    .000578    -3.29   0.001    -.0030347   -.0007688
                       |
                   age |  -.0018405   .0002832    -6.50   0.000    -.0023955   -.0012855
                female |  -.0114943   .0074259    -1.55   0.122    -.0260489    .0030602
                unempl |   -.006177   .0082751    -0.75   0.455     -.022396    .0100419
                 urban |   .0464895    .009199     5.05   0.000     .0284598    .0645193
              interest |  -.0308277   .0035831    -8.60   0.000    -.0378505   -.0238049
                  news |    .008149   .0027905     2.92   0.003     .0026797    .0136184
                       |
                 round |
                    5  |   .0341397   .0109838     3.11   0.002      .012612    .0556675
                    6  |  -.0354053   .0110204    -3.21   0.001    -.0570048   -.0138057
                       |
                 _cons |    .542573   .0285361    19.01   0.000     .4866432    .5985027
-----------------------+----------------------------------------------------------------
               sigma_u |  .07530807
               sigma_e |  .47806341
                   rho |  .02421401   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     17,912
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0182                                         min =        197
     Between = 0.1408                                         avg =      995.1
     Overall = 0.0209                                         max =      2,258

                                                Wald chi2(11)     =     335.20
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0072191   .0028277     2.55   0.011     .0016769    .0127613
               educlvl |   .0173515   .0037098     4.68   0.000     .0100805    .0246225
                       |
c.lnfdi_jobs#c.educlvl |  -.0011118   .0005909    -1.88   0.060      -.00227    .0000464
                       |
                   age |  -.0015124   .0002831    -5.34   0.000    -.0020674   -.0009574
                female |  -.0080746   .0073892    -1.09   0.274    -.0225572    .0064079
                unempl |  -.0079428   .0081937    -0.97   0.332    -.0240021    .0081165
                 urban |   .0551853   .0091255     6.05   0.000     .0372995     .073071
              interest |  -.0309063   .0035437    -8.72   0.000    -.0378518   -.0239608
                  news |   .0069638   .0027894     2.50   0.013     .0014966     .012431
                       |
                 round |
                    5  |   .0311566   .0108994     2.86   0.004     .0097942     .052519
                    6  |  -.0372032   .0109433    -3.40   0.001    -.0586516   -.0157548
                       |
                 _cons |   .5477562    .029138    18.80   0.000     .4906468    .6048656
-----------------------+----------------------------------------------------------------
               sigma_u |  .08020931
               sigma_e |   .4764631
                   rho |  .02755843   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     16,325
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0177                                         min =        197
     Between = 0.2293                                         avg =      906.9
     Overall = 0.0239                                         max =      2,002

                                                Wald chi2(11)     =     299.05
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0133088   .0028609     4.65   0.000     .0077016     .018916
               educlvl |   .0203339   .0039267     5.18   0.000     .0126377    .0280301
                       |
c.lnfdi_jobs#c.educlvl |  -.0021727   .0005932    -3.66   0.000    -.0033353   -.0010101
                       |
                   age |  -.0017424   .0002918    -5.97   0.000    -.0023143   -.0011704
                female |   .0007635   .0077246     0.10   0.921    -.0143764    .0159034
                unempl |  -.0022357   .0084777    -0.26   0.792    -.0188517    .0143802
                 urban |   .0461407   .0098141     4.70   0.000     .0269053     .065376
              interest |  -.0302735    .003718    -8.14   0.000    -.0375606   -.0229865
                  news |   .0076196   .0028893     2.64   0.008     .0019566    .0132825
                       |
                 round |
                    5  |   .0468573    .012291     3.81   0.000     .0227673    .0709472
                    6  |  -.0268276   .0122537    -2.19   0.029    -.0508445   -.0028107
                       |
                 _cons |   .5139514   .0306044    16.79   0.000     .4539678     .573935
-----------------------+----------------------------------------------------------------
               sigma_u |  .08174452
               sigma_e |   .4756044
                   rho |  .02869339   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     16,892
Group variable: cname                           Number of groups  =         18

R-squared:                                      Obs per group:
     Within  = 0.0174                                         min =        197
     Between = 0.2186                                         avg =      938.4
     Overall = 0.0222                                         max =      2,258

                                                Wald chi2(11)     =     304.92
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
             trustelec | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0107535   .0027994     3.84   0.000     .0052668    .0162402
               educlvl |   .0199728   .0037713     5.30   0.000     .0125812    .0273645
                       |
c.lnfdi_jobs#c.educlvl |  -.0019557   .0005828    -3.36   0.001    -.0030979   -.0008135
                       |
                   age |  -.0019463   .0002915    -6.68   0.000    -.0025176   -.0013751
                female |  -.0071381   .0076393    -0.93   0.350    -.0221108    .0078347
                unempl |  -.0001895   .0084381    -0.02   0.982    -.0167279    .0163489
                 urban |   .0437421   .0093887     4.66   0.000     .0253405    .0621437
              interest |   -.032866   .0036512    -9.00   0.000    -.0400221   -.0257098
                  news |    .004758   .0028689     1.66   0.097    -.0008649    .0103809
                       |
                 round |
                    5  |   .0415251    .011617     3.57   0.000     .0187562    .0642941
                    6  |  -.0264313   .0117333    -2.25   0.024    -.0494282   -.0034343
                       |
                 _cons |   .5441244   .0282816    19.24   0.000     .4886935    .5995552
-----------------------+----------------------------------------------------------------
               sigma_u |  .07124361
               sigma_e |  .47813037
                   rho |  .02172012   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. coefplot trustelec_1 trustelec_4 trustelec_6 trustelec_8 trustelec_9 ///
>         trustelec_11 trustelec_12 trustelec_16 trustelec_18 trustelec_20 ///
>         trustelec_21 trustelec_22 trustelec_24 trustelec_29 trustelec_30 ///
>         trustelec_31 trustelec_32 trustelec_33 trustelec_34 trustelec_36, ///
>         keep(lnfdi_jobs c.lnfdi_jobs#c.educlvl) xline(0) ///
>         title("Distrust in Electoral Commission") ///
>         rename(lnfdi_jobs = "FDI exposure" c.lnfdi_jobs#c.educlvl = "FDI * education") ///
>         legend(order(- "Country excluded:" " " 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40) ///
>                 label(2 "Algeria") label(4 "Burkina Faso") label(6 "Cameroon") label(8 "Egypt") label(10 "Gabon") 
> ///
>                 label(12 "Guinea") label(14 "Ivory Coast") label(16 "Madagascar") label(18 "Mali") label(20 "Moroc
> co") ///
>                 label(22 "Mozambique") label(24 "Namibia") label(26 "Nigeria") label(28 "Sudan") label(30 "Swazila
> nd") ///
>                 label(32 "Tanzania") label(34 "Togo") label(36 "Tunisia") label(38 "Uganda") label(40 "Zimbabwe"))
>  ///
>         name(afb_jackknife2, replace)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)

. * Leave out one country test: Opinion: Country Is Not a Democracy
. foreach i in 1 4 6 8 9 11 12 16 18 20 21 22 24 29 30 31 32 33 34 36 {
  2.         xtreg demoqua c.lnfdi_jobs##c.educlvl $C1 $C2 i.round if cname!=`i', i(cname)
  3.         est store demoqua_`i'
  4. }

Random-effects GLS regression                   Number of obs     =     19,257
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0191                                         min =        205
     Between = 0.2184                                         avg =    1,013.5
     Overall = 0.0274                                         max =      2,130

                                                Wald chi2(11)     =     379.10
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0164922   .0048374     3.41   0.001      .007011    .0259734
               educlvl |   .0550645    .006579     8.37   0.000       .04217    .0679591
                       |
c.lnfdi_jobs#c.educlvl |  -.0041214   .0009828    -4.19   0.000    -.0060476   -.0021953
                       |
                   age |  -.0002006   .0004924    -0.41   0.684    -.0011656    .0007644
                female |  -.0569097    .012866    -4.42   0.000    -.0821265   -.0316929
                unempl |  -.0316435   .0141834    -2.23   0.026    -.0594425   -.0038446
                 urban |   .0696363   .0160701     4.33   0.000     .0381396     .101133
              interest |  -.0493932   .0062545    -7.90   0.000    -.0616518   -.0371345
                  news |   .0101834   .0048353     2.11   0.035     .0007065    .0196604
                       |
                 round |
                    5  |   .1639752   .0193568     8.47   0.000     .1260367    .2019137
                    6  |   .0659396   .0199432     3.31   0.001     .0268516    .1050276
                       |
                 _cons |   2.397804   .0583364    41.10   0.000     2.283467    2.512142
-----------------------+----------------------------------------------------------------
               sigma_u |  .18586106
               sigma_e |  .86005602
                   rho |  .04461705   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,641
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0176                                         min =        205
     Between = 0.1834                                         avg =    1,033.7
     Overall = 0.0246                                         max =      2,130

                                                Wald chi2(11)     =     354.12
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0152707   .0048217     3.17   0.002     .0058204    .0247209
               educlvl |   .0480624   .0066815     7.19   0.000     .0349668     .061158
                       |
c.lnfdi_jobs#c.educlvl |  -.0035109    .000981    -3.58   0.000    -.0054336   -.0015882
                       |
                   age |  -.0004748    .000486    -0.98   0.329    -.0014274    .0004777
                female |  -.0623923   .0127296    -4.90   0.000    -.0873419   -.0374427
                unempl |  -.0338641   .0140088    -2.42   0.016    -.0613208   -.0064075
                 urban |   .0680759   .0157419     4.32   0.000     .0372224    .0989294
              interest |  -.0501261   .0062269    -8.05   0.000    -.0623305   -.0379217
                  news |   .0104857   .0047386     2.21   0.027     .0011983    .0197732
                       |
                 round |
                    5  |   .1581063   .0196189     8.06   0.000      .119654    .1965585
                    6  |   .0594751   .0201899     2.95   0.003     .0199036    .0990465
                       |
                 _cons |   2.432206   .0598405    40.64   0.000     2.314921    2.549491
-----------------------+----------------------------------------------------------------
               sigma_u |  .19312057
               sigma_e |  .85881978
                   rho |  .04813156   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,513
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0184                                         min =        205
     Between = 0.2507                                         avg =    1,027.0
     Overall = 0.0268                                         max =      2,130

                                                Wald chi2(11)     =     372.72
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0170615   .0048005     3.55   0.000     .0076526    .0264704
               educlvl |   .0541831   .0066362     8.16   0.000     .0411764    .0671897
                       |
c.lnfdi_jobs#c.educlvl |   -.004148   .0009803    -4.23   0.000    -.0060693   -.0022268
                       |
                   age |  -.0004203   .0004891    -0.86   0.390    -.0013789    .0005383
                female |  -.0652576   .0128517    -5.08   0.000    -.0904465   -.0400686
                unempl |  -.0245582   .0141203    -1.74   0.082    -.0522334    .0031171
                 urban |   .0703549   .0158998     4.42   0.000     .0391919    .1015179
              interest |  -.0506605   .0062871    -8.06   0.000    -.0629829    -.038338
                  news |   .0103082   .0047859     2.15   0.031      .000928    .0196885
                       |
                 round |
                    5  |   .1596374   .0194077     8.23   0.000     .1215991    .1976758
                    6  |   .0739881   .0200002     3.70   0.000     .0347885    .1131877
                       |
                 _cons |   2.403663   .0511799    46.96   0.000     2.303352    2.503974
-----------------------+----------------------------------------------------------------
               sigma_u |  .14023259
               sigma_e |  .86370025
                   rho |  .02568452   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,183
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0190                                         min =        205
     Between = 0.2464                                         avg =    1,009.6
     Overall = 0.0286                                         max =      2,130

                                                Wald chi2(11)     =     376.25
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0168373   .0050544     3.33   0.001     .0069309    .0267438
               educlvl |   .0546095   .0068814     7.94   0.000     .0411223    .0680967
                       |
c.lnfdi_jobs#c.educlvl |  -.0038219   .0010387    -3.68   0.000    -.0058578    -.001786
                       |
                   age |  -.0007321   .0004913    -1.49   0.136    -.0016951    .0002309
                female |  -.0631606   .0128025    -4.93   0.000    -.0882531   -.0380681
                unempl |  -.0313415   .0141287    -2.22   0.027    -.0590333   -.0036497
                 urban |   .0618518   .0161454     3.83   0.000     .0302073    .0934963
              interest |  -.0504747    .006205    -8.13   0.000    -.0626364   -.0383131
                  news |   .0125738   .0047914     2.62   0.009     .0031827    .0219649
                       |
                 round |
                    5  |   .1440838   .0192909     7.47   0.000     .1062744    .1818932
                    6  |   .0813957   .0199182     4.09   0.000     .0423568    .1204346
                       |
                 _cons |   2.395534   .0613799    39.03   0.000     2.275231    2.515836
-----------------------+----------------------------------------------------------------
               sigma_u |  .20069166
               sigma_e |  .85632971
                   rho |  .05206606   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,452
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0190                                         min =        205
     Between = 0.1614                                         avg =    1,023.8
     Overall = 0.0251                                         max =      2,130

                                                Wald chi2(11)     =     379.77
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0180017   .0047616     3.78   0.000     .0086691    .0273342
               educlvl |    .054101   .0065438     8.27   0.000     .0412753    .0669267
                       |
c.lnfdi_jobs#c.educlvl |  -.0040624   .0009683    -4.20   0.000    -.0059602   -.0021647
                       |
                   age |  -.0004655   .0004918    -0.95   0.344    -.0014294    .0004984
                female |  -.0695838   .0128744    -5.40   0.000    -.0948172   -.0443504
                unempl |  -.0331111   .0142036    -2.33   0.020    -.0609497   -.0052725
                 urban |   .0681037   .0158013     4.31   0.000     .0371338    .0990736
              interest |  -.0497993   .0063004    -7.90   0.000    -.0621478   -.0374508
                  news |   .0080436   .0048014     1.68   0.094    -.0013669    .0174541
                       |
                 round |
                    5  |   .1567597   .0193785     8.09   0.000     .1187784    .1947409
                    6  |   .0666632   .0199685     3.34   0.001     .0275256    .1058007
                       |
                 _cons |   2.386831   .0607349    39.30   0.000     2.267793    2.505869
-----------------------+----------------------------------------------------------------
               sigma_u |  .20038888
               sigma_e |  .86471711
                   rho |  .05096604   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,837
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0195                                         min =        205
     Between = 0.1895                                         avg =    1,044.1
     Overall = 0.0270                                         max =      2,130

                                                Wald chi2(11)     =     398.50
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0184376   .0047399     3.89   0.000     .0091475    .0277276
               educlvl |   .0554988   .0065244     8.51   0.000     .0427112    .0682863
                       |
c.lnfdi_jobs#c.educlvl |  -.0042137   .0009633    -4.37   0.000    -.0061017   -.0023256
                       |
                   age |  -.0004143   .0004874    -0.85   0.395    -.0013696     .000541
                female |  -.0661442    .012695    -5.21   0.000    -.0910258   -.0412625
                unempl |  -.0362884   .0139738    -2.60   0.009    -.0636765   -.0089004
                 urban |   .0689917   .0157463     4.38   0.000     .0381295    .0998538
              interest |  -.0523265   .0062303    -8.40   0.000    -.0645376   -.0401154
                  news |   .0097229   .0047325     2.05   0.040     .0004474    .0189983
                       |
                 round |
                    5  |   .1587567   .0193268     8.21   0.000     .1208769    .1966366
                    6  |   .0654728   .0199173     3.29   0.001     .0264356      .10451
                       |
                 _cons |   2.397132   .0590769    40.58   0.000     2.281344    2.512921
-----------------------+----------------------------------------------------------------
               sigma_u |   .1919798
               sigma_e |  .86161138
                   rho |  .04729826   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,377
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0180                                         min =        205
     Between = 0.2044                                         avg =    1,019.8
     Overall = 0.0256                                         max =      2,130

                                                Wald chi2(11)     =     359.31
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0193249   .0048392     3.99   0.000     .0098402    .0288096
               educlvl |   .0516243   .0065651     7.86   0.000     .0387568    .0644917
                       |
c.lnfdi_jobs#c.educlvl |  -.0044199   .0009775    -4.52   0.000    -.0063358    -.002504
                       |
                   age |  -.0006193   .0004881    -1.27   0.205     -.001576    .0003374
                female |  -.0678435   .0128012    -5.30   0.000    -.0929334   -.0427536
                unempl |  -.0348517   .0141161    -2.47   0.014    -.0625188   -.0071846
                 urban |   .0789771   .0158201     4.99   0.000     .0479704    .1099839
              interest |  -.0467667   .0062549    -7.48   0.000     -.059026   -.0345074
                  news |   .0091617   .0047978     1.91   0.056    -.0002418    .0185651
                       |
                 round |
                    5  |   .1572004   .0193148     8.14   0.000     .1193441    .1950566
                    6  |   .0705631   .0199193     3.54   0.000      .031522    .1096043
                       |
                 _cons |   2.412141   .0578885    41.67   0.000     2.298682    2.525601
-----------------------+----------------------------------------------------------------
               sigma_u |  .18352205
               sigma_e |  .85876525
                   rho |  .04367501   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,422
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0181                                         min =        205
     Between = 0.1790                                         avg =    1,022.2
     Overall = 0.0244                                         max =      2,130

                                                Wald chi2(11)     =     360.28
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0166486   .0047507     3.50   0.000     .0073374    .0259598
               educlvl |   .0518255   .0065303     7.94   0.000     .0390263    .0646246
                       |
c.lnfdi_jobs#c.educlvl |  -.0038622   .0009634    -4.01   0.000    -.0057503    -.001974
                       |
                   age |  -.0006332   .0004895    -1.29   0.196    -.0015926    .0003263
                female |  -.0664716   .0127672    -5.21   0.000    -.0914949   -.0414483
                unempl |  -.0348584   .0140288    -2.48   0.013    -.0623543   -.0073625
                 urban |   .0841375   .0160878     5.23   0.000     .0526061     .115669
              interest |  -.0495125   .0062255    -7.95   0.000    -.0617143   -.0373106
                  news |    .005539   .0047797     1.16   0.247     -.003829     .014907
                       |
                 round |
                    5  |   .1199762    .019823     6.05   0.000     .0811239    .1588286
                    6  |   .0185973   .0204541     0.91   0.363     -.021492    .0586866
                       |
                 _cons |   2.451315   .0609621    40.21   0.000     2.331832    2.570799
-----------------------+----------------------------------------------------------------
               sigma_u |  .20119761
               sigma_e |  .85617478
                   rho |   .0523331   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,558
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0182                                         min =        205
     Between = 0.2008                                         avg =    1,029.4
     Overall = 0.0264                                         max =      2,130

                                                Wald chi2(11)     =     366.14
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0203169   .0048605     4.18   0.000     .0107905    .0298433
               educlvl |   .0564271   .0067576     8.35   0.000     .0431823    .0696718
                       |
c.lnfdi_jobs#c.educlvl |  -.0045505   .0009901    -4.60   0.000    -.0064911   -.0026099
                       |
                   age |  -.0006575   .0004906    -1.34   0.180    -.0016189     .000304
                female |  -.0647508   .0128116    -5.05   0.000     -.089861   -.0396405
                unempl |  -.0293475   .0141249    -2.08   0.038    -.0570319   -.0016632
                 urban |   .0672539   .0159131     4.23   0.000     .0360647     .098443
              interest |  -.0500393   .0062745    -7.98   0.000    -.0623371   -.0377415
                  news |   .0098003   .0047692     2.05   0.040     .0004529    .0191476
                       |
                 round |
                    5  |   .1528302   .0196766     7.77   0.000     .1142648    .1913955
                    6  |    .071332   .0202155     3.53   0.000     .0317104    .1109537
                       |
                 _cons |   2.399398   .0608184    39.45   0.000     2.280196      2.5186
-----------------------+----------------------------------------------------------------
               sigma_u |  .19733556
               sigma_e |  .86326043
                   rho |  .04965993   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,048
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0194                                         min =        205
     Between = 0.1942                                         avg =    1,002.5
     Overall = 0.0271                                         max =      2,130

                                                Wald chi2(11)     =     380.62
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0224706   .0049546     4.54   0.000     .0127598    .0321814
               educlvl |   .0533671   .0066351     8.04   0.000     .0403626    .0663716
                       |
c.lnfdi_jobs#c.educlvl |  -.0040992   .0010175    -4.03   0.000    -.0060935   -.0021049
                       |
                   age |  -.0002231   .0004929    -0.45   0.651    -.0011891     .000743
                female |  -.0621329   .0129311    -4.80   0.000    -.0874774   -.0367884
                unempl |  -.0237957   .0142751    -1.67   0.096    -.0517745     .004183
                 urban |   .0687427    .016054     4.28   0.000     .0372774    .1002081
              interest |  -.0539815   .0062663    -8.61   0.000    -.0662632   -.0416998
                  news |   .0098988   .0048377     2.05   0.041     .0004172    .0193804
                       |
                 round |
                    5  |   .1491899   .0193739     7.70   0.000     .1112179     .187162
                    6  |   .0570247   .0199727     2.86   0.004      .017879    .0961704
                       |
                 _cons |   2.381919   .0609885    39.06   0.000     2.262384    2.501454
-----------------------+----------------------------------------------------------------
               sigma_u |  .20113661
               sigma_e |  .85983502
                   rho |  .05188174   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,739
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0193                                         min =        205
     Between = 0.1829                                         avg =      986.3
     Overall = 0.0269                                         max =      2,130

                                                Wald chi2(11)     =     372.21
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .018743   .0049553     3.78   0.000     .0090308    .0284551
               educlvl |   .0566454   .0066769     8.48   0.000     .0435589     .069732
                       |
c.lnfdi_jobs#c.educlvl |  -.0044589   .0009908    -4.50   0.000    -.0064007    -.002517
                       |
                   age |  -.0005988   .0004978    -1.20   0.229    -.0015745     .000377
                female |   -.067467   .0131027    -5.15   0.000    -.0931479   -.0417861
                unempl |  -.0335821   .0143399    -2.34   0.019    -.0616878   -.0054765
                 urban |   .0614885   .0162115     3.79   0.000     .0297146    .0932624
              interest |   -.050115   .0064441    -7.78   0.000    -.0627452   -.0374848
                  news |   .0085734   .0048928     1.75   0.080    -.0010164    .0181632
                       |
                 round |
                    5  |   .1715036   .0203113     8.44   0.000     .1316943     .211313
                    6  |    .079979   .0209002     3.83   0.000     .0390153    .1209428
                       |
                 _cons |   2.414898   .0603158    40.04   0.000     2.296681    2.533114
-----------------------+----------------------------------------------------------------
               sigma_u |  .19240065
               sigma_e |  .86276628
                   rho |  .04737495   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,779
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0194                                         min =        205
     Between = 0.2781                                         avg =      988.4
     Overall = 0.0300                                         max =      2,130

                                                Wald chi2(11)     =     376.95
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0203325   .0048709     4.17   0.000     .0107858    .0298792
               educlvl |   .0555697   .0066551     8.35   0.000      .042526    .0686134
                       |
c.lnfdi_jobs#c.educlvl |  -.0042172   .0009868    -4.27   0.000    -.0061513   -.0022832
                       |
                   age |  -.0004104   .0005007    -0.82   0.412    -.0013918    .0005711
                female |  -.0631612   .0131366    -4.81   0.000    -.0889084   -.0374139
                unempl |  -.0328661   .0142427    -2.31   0.021    -.0607814   -.0049508
                 urban |   .0659059   .0161995     4.07   0.000     .0341555    .0976564
              interest |  -.0484006   .0064233    -7.54   0.000    -.0609901   -.0358112
                  news |   .0119003   .0048525     2.45   0.014     .0023895    .0214111
                       |
                 round |
                    5  |   .1597366   .0206298     7.74   0.000     .1193029    .2001704
                    6  |   .0713825   .0212865     3.35   0.001     .0296617    .1131034
                       |
                 _cons |   2.412126    .057115    42.23   0.000     2.300183    2.524069
-----------------------+----------------------------------------------------------------
               sigma_u |  .17438266
               sigma_e |  .86445584
                   rho |  .03910194   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,153
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0212                                         min =        205
     Between = 0.1812                                         avg =      955.4
     Overall = 0.0284                                         max =      2,130

                                                Wald chi2(11)     =     396.31
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |    .019875   .0049528     4.01   0.000     .0101676    .0295823
               educlvl |   .0572835   .0068774     8.33   0.000      .043804     .070763
                       |
c.lnfdi_jobs#c.educlvl |    -.00445   .0010197    -4.36   0.000    -.0064486   -.0024515
                       |
                   age |   -.000334    .000505    -0.66   0.508    -.0013238    .0006559
                female |  -.0680119   .0133015    -5.11   0.000    -.0940823   -.0419415
                unempl |  -.0321295   .0145865    -2.20   0.028    -.0607185   -.0035405
                 urban |    .072012   .0166104     4.34   0.000     .0394562    .1045678
              interest |  -.0474785    .006536    -7.26   0.000    -.0602889   -.0346681
                  news |   .0118487   .0049711     2.38   0.017     .0021054    .0215919
                       |
                 round |
                    5  |   .1715949   .0214984     7.98   0.000     .1294588     .213731
                    6  |   .0591376   .0219894     2.69   0.007     .0160391     .102236
                       |
                 _cons |   2.370969   .0610336    38.85   0.000     2.251345    2.490592
-----------------------+----------------------------------------------------------------
               sigma_u |  .19521255
               sigma_e |  .86367869
                   rho |  .04860404   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,934
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0188                                         min =        302
     Between = 0.1336                                         avg =    1,049.2
     Overall = 0.0259                                         max =      2,130

                                                Wald chi2(11)     =     383.89
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0187621   .0047404     3.96   0.000      .009471    .0280532
               educlvl |   .0546564   .0064792     8.44   0.000     .0419573    .0673554
                       |
c.lnfdi_jobs#c.educlvl |  -.0042663   .0009645    -4.42   0.000    -.0061567    -.002376
                       |
                   age |  -.0004359   .0004834    -0.90   0.367    -.0013834    .0005115
                female |  -.0652221   .0126645    -5.15   0.000    -.0900441      -.0404
                unempl |  -.0295894   .0139179    -2.13   0.034    -.0568679   -.0023109
                 urban |   .0708422   .0157165     4.51   0.000     .0400384    .1016459
              interest |  -.0511346   .0061731    -8.28   0.000    -.0632337   -.0390355
                  news |   .0080762   .0047305     1.71   0.088    -.0011954    .0173478
                       |
                 round |
                    5  |   .1572851    .019309     8.15   0.000     .1194401    .1951301
                    6  |   .0654394   .0198904     3.29   0.001     .0264548    .1044239
                       |
                 _cons |   2.381326   .0604307    39.41   0.000     2.262884    2.499768
-----------------------+----------------------------------------------------------------
               sigma_u |  .20074115
               sigma_e |  .86119239
                   rho |  .05153404   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,363
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0178                                         min =        205
     Between = 0.1785                                         avg =    1,019.1
     Overall = 0.0246                                         max =      2,130

                                                Wald chi2(11)     =     354.53
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0179468   .0047363     3.79   0.000     .0086638    .0272297
               educlvl |   .0501425   .0065425     7.66   0.000     .0373195    .0629656
                       |
c.lnfdi_jobs#c.educlvl |  -.0037414   .0009644    -3.88   0.000    -.0056317   -.0018512
                       |
                   age |  -.0001786     .00049    -0.36   0.715    -.0011391    .0007818
                female |  -.0649787   .0128107    -5.07   0.000    -.0900872   -.0398701
                unempl |  -.0316162   .0140373    -2.25   0.024    -.0591289   -.0041035
                 urban |   .0638431   .0161235     3.96   0.000     .0322417    .0954446
              interest |  -.0510621   .0062654    -8.15   0.000    -.0633421    -.038782
                  news |   .0072892   .0047656     1.53   0.126    -.0020513    .0166296
                       |
                 round |
                    5  |   .1600191   .0192701     8.30   0.000     .1222504    .1977879
                    6  |      .0614   .0198795     3.09   0.002      .022437    .1003631
                       |
                 _cons |   2.398374   .0536914    44.67   0.000      2.29314    2.503607
-----------------------+----------------------------------------------------------------
               sigma_u |  .15830669
               sigma_e |  .85758952
                   rho |  .03295243   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,758
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0183                                         min =        205
     Between = 0.1719                                         avg =      987.3
     Overall = 0.0247                                         max =      2,130

                                                Wald chi2(11)     =     353.11
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0158043   .0049232     3.21   0.001     .0061551    .0254536
               educlvl |   .0525527    .006642     7.91   0.000     .0395346    .0655709
                       |
c.lnfdi_jobs#c.educlvl |  -.0038101   .0009861    -3.86   0.000    -.0057427   -.0018775
                       |
                   age |  -.0003626   .0005059    -0.72   0.474     -.001354    .0006289
                female |  -.0641036   .0131545    -4.87   0.000    -.0898859   -.0383213
                unempl |    -.03136   .0144277    -2.17   0.030    -.0596378   -.0030822
                 urban |   .0729583    .016455     4.43   0.000      .040707    .1052096
              interest |  -.0524649   .0064131    -8.18   0.000    -.0650344   -.0398953
                  news |   .0081838   .0049356     1.66   0.097    -.0014898    .0178575
                       |
                 round |
                    5  |   .1531028   .0202601     7.56   0.000     .1133936    .1928119
                    6  |   .0657884    .020781     3.17   0.002     .0250584    .1065184
                       |
                 _cons |   2.445175   .0601424    40.66   0.000     2.327298    2.563052
-----------------------+----------------------------------------------------------------
               sigma_u |  .19212446
               sigma_e |  .86811759
                   rho |  .04669189   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     19,344
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0198                                         min =        205
     Between = 0.1943                                         avg =    1,018.1
     Overall = 0.0274                                         max =      2,130

                                                Wald chi2(11)     =     394.41
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0193604   .0048077     4.03   0.000     .0099374    .0287834
               educlvl |   .0597898   .0066671     8.97   0.000     .0467224    .0728571
                       |
c.lnfdi_jobs#c.educlvl |  -.0046831   .0009784    -4.79   0.000    -.0066007   -.0027654
                       |
                   age |  -.0007441   .0004918    -1.51   0.130     -.001708    .0002198
                female |  -.0688844   .0128596    -5.36   0.000    -.0940888   -.0436799
                unempl |   -.031803   .0142067    -2.24   0.025    -.0596477   -.0039584
                 urban |   .0599223   .0159829     3.75   0.000     .0285963    .0912482
              interest |  -.0517541   .0063105    -8.20   0.000    -.0641225   -.0393857
                  news |   .0087725    .004779     1.84   0.066    -.0005942    .0181393
                       |
                 round |
                    5  |   .1588422   .0193786     8.20   0.000     .1208607    .1968236
                    6  |   .0690314   .0199837     3.45   0.001     .0298642    .1081987
                       |
                 _cons |   2.396928   .0605012    39.62   0.000     2.278348    2.515508
-----------------------+----------------------------------------------------------------
               sigma_u |   .1976328
               sigma_e |  .86187357
                   rho |  .04995458   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,758
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0178                                         min =        205
     Between = 0.2341                                         avg =      987.3
     Overall = 0.0277                                         max =      2,130

                                                Wald chi2(11)     =     344.42
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0164776   .0049525     3.33   0.001      .006771    .0261842
               educlvl |   .0529176   .0065927     8.03   0.000     .0399962    .0658391
                       |
c.lnfdi_jobs#c.educlvl |  -.0037665   .0010111    -3.73   0.000    -.0057482   -.0017849
                       |
                   age |  -.0001334   .0005031    -0.27   0.791    -.0011195    .0008527
                female |  -.0645813   .0130225    -4.96   0.000    -.0901049   -.0390577
                unempl |  -.0288545   .0143738    -2.01   0.045    -.0570265   -.0006824
                 urban |   .0681804   .0161774     4.21   0.000     .0364732    .0998876
              interest |  -.0512047   .0063304    -8.09   0.000    -.0636121   -.0387973
                  news |   .0075231   .0048831     1.54   0.123    -.0020477    .0170938
                       |
                 round |
                    5  |   .1472813   .0193868     7.60   0.000     .1092838    .1852788
                    6  |   .0771153   .0200162     3.85   0.000     .0378842    .1163464
                       |
                 _cons |   2.393809   .0607758    39.39   0.000      2.27469    2.512927
-----------------------+----------------------------------------------------------------
               sigma_u |  .19858383
               sigma_e |  .86008369
                   rho |  .05061162   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,009
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0176                                         min =        205
     Between = 0.1814                                         avg =      947.8
     Overall = 0.0236                                         max =      1,986

                                                Wald chi2(11)     =     325.77
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0193675   .0049542     3.91   0.000     .0096575    .0290775
               educlvl |   .0541684   .0069385     7.81   0.000     .0405692    .0677675
                       |
c.lnfdi_jobs#c.educlvl |  -.0042744   .0010057    -4.25   0.000    -.0062456   -.0023032
                       |
                   age |  -.0004146   .0005051    -0.82   0.412    -.0014045    .0005754
                female |  -.0545665   .0133438    -4.09   0.000    -.0807198   -.0284131
                unempl |  -.0296669   .0145228    -2.04   0.041    -.0581311   -.0012028
                 urban |   .0634827   .0169499     3.75   0.000     .0302615    .0967039
              interest |  -.0491591   .0065339    -7.52   0.000    -.0619653   -.0363529
                  news |   .0098787   .0049396     2.00   0.046     .0001972    .0195602
                       |
                 round |
                    5  |   .1795086   .0217451     8.26   0.000      .136889    .2221282
                    6  |   .0920982   .0222161     4.15   0.000     .0485555    .1356408
                       |
                 _cons |   2.381354   .0583312    40.82   0.000     2.267027    2.495681
-----------------------+----------------------------------------------------------------
               sigma_u |  .17346741
               sigma_e |  .86203344
                   rho |  .03891779   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

Random-effects GLS regression                   Number of obs     =     18,516
Group variable: cname                           Number of groups  =         19

R-squared:                                      Obs per group:
     Within  = 0.0178                                         min =        205
     Between = 0.2206                                         avg =      974.5
     Overall = 0.0266                                         max =      2,130

                                                Wald chi2(11)     =     340.88
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               demoqua | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0177268   .0047874     3.70   0.000     .0083438    .0271098
               educlvl |   .0510436   .0065865     7.75   0.000     .0381344    .0639529
                       |
c.lnfdi_jobs#c.educlvl |  -.0039326   .0009732    -4.04   0.000    -.0058401   -.0020252
                       |
                   age |  -.0006959   .0005002    -1.39   0.164    -.0016762    .0002845
                female |  -.0645004   .0130859    -4.93   0.000    -.0901482   -.0388525
                unempl |  -.0318144   .0143215    -2.22   0.026     -.059884   -.0037447
                 urban |   .0650308    .016154     4.03   0.000     .0333695    .0966922
              interest |  -.0510995   .0063667    -8.03   0.000    -.0635781    -.038621
                  news |   .0070738   .0048597     1.46   0.146    -.0024511    .0165987
                       |
                 round |
                    5  |    .162974   .0203625     8.00   0.000     .1230642    .2028838
                    6  |   .0913975   .0210893     4.33   0.000     .0500632    .1327318
                       |
                 _cons |   2.401827   .0567519    42.32   0.000     2.290595    2.513058
-----------------------+----------------------------------------------------------------
               sigma_u |  .17391623
               sigma_e |   .8568778
                   rho |  .03956494   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. coefplot demoqua_1 demoqua_4 demoqua_6 demoqua_8 demoqua_9 ///
>         demoqua_11 demoqua_12 demoqua_16 demoqua_18 demoqua_20 ///
>         demoqua_21 demoqua_22 demoqua_24 demoqua_29 demoqua_30 ///
>         demoqua_31 demoqua_32 demoqua_33 demoqua_34 demoqua_36, ///
>         keep(lnfdi_jobs c.lnfdi_jobs#c.educlvl) xline(0) ///
>         title("Opinion: Country Is Not a Democracy") ///
>         rename(lnfdi_jobs = "FDI exposure" c.lnfdi_jobs#c.educlvl = "FDI * education") ///
>         legend(order(- "Country excluded:" " " 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40) ///
>                 label(2 "Algeria") label(4 "Burkina Faso") label(6 "Cameroon") label(8 "Egypt") label(10 "Gabon") 
> ///
>                 label(12 "Guinea") label(14 "Ivory Coast") label(16 "Madagascar") label(18 "Mali") label(20 "Moroc
> co") ///
>                 label(22 "Mozambique") label(24 "Namibia") label(26 "Nigeria") label(28 "Sudan") label(30 "Swazila
> nd") ///
>                 label(32 "Tanzania") label(34 "Togo") label(36 "Tunisia") label(38 "Uganda") label(40 "Zimbabwe"))
>  ///
>         name(afb_jackknife3, replace)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)

. 
. *** FIGURE A-4: Combined jackkknife graph
. graph combine afb_jackknife1 afb_jackknife2 afb_jackknife3, col(1) iscale(.5) ///
>         xsize(5) ysize(6) name(afb_jackknife, replace)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p16 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p17 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p18 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p19 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)
(note:  named style p20 not found in class seriesstyle, default attributes used)

. 
. 
. 
. ***** 2. Pew Global Attitudes Survey Data
. *------------------------------------------------------------------------------*
. 
. *** Load data
. use "Rommel_FDIsupport_CPS_PEW.dta", clear

. 
. *** Model specification
. * Controls
. global C1 age gender unemp urban

. global C2 income intnews 

. global C3 religion remit married children relatives 

. * Sample
. xtreg problem c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,646
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0134                                         min =        446
     Between = 0.0692                                         avg =      760.4
     Overall = 0.0145                                         max =      1,029

                                                Wald chi2(9)      =     145.18
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0059443   .0032836     1.81   0.070    -.0004915      .01238
                educlvl |   .0099307   .0060936     1.63   0.103    -.0020126     .021874
                        |
c.lnfdi_capex#c.educlvl |  -.0025876   .0009506    -2.72   0.006    -.0044507   -.0007244
                        |
                    age |  -.0015805     .00039    -4.05   0.000    -.0023449   -.0008161
                 gender |  -.0139884    .009899    -1.41   0.158      -.03339    .0054132
                  unemp |   .0252284   .0105306     2.40   0.017     .0045888     .045868
                  urban |   .0511622    .010663     4.80   0.000     .0302631    .0720613
                 income |  -.0246401   .0026849    -9.18   0.000    -.0299025   -.0193777
                intnews |    .024564    .009193     2.67   0.008      .006546    .0425819
                  _cons |   .3219699   .0345449     9.32   0.000     .2542631    .3896767
------------------------+----------------------------------------------------------------
                sigma_u |  .07879719
                sigma_e |   .4605604
                    rho |  .02843925   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. gen sample=e(sample)==1

. 
. *** TABLE A-12: Summary statistics, Pew Global Attitudes
. tabstat problem worse leadsat legit_free legit_judi legit_mili ///
>         educlvl lnfdi_capex lnfdi_jobs age gender unemp urban income intnews ///
>         religion remit married children relatives if sample==1, ///
>         stats(N mean sd min max) format(%10.2f)

   Stats |    problem      worse    leadsat   legit_~e   legit~di   legit~li    educlvl   lnfdi_~x   lnfdi_~s
---------+---------------------------------------------------------------------------------------------------
       N |   10646.00    9602.00    9992.00   10411.00   10337.00    9806.00   10646.00   10646.00   10646.00
    Mean |       0.33      -1.35       1.97       0.86       1.17       0.71       2.95       5.25       6.27
      SD |       0.47       2.07       1.00       0.99       1.10       0.96       1.55       3.17       3.58
     Min |       0.00     -10.00       1.00       0.00       0.00       0.00       0.00       0.00       0.00
     Max |       1.00      10.00       4.00       3.00       3.00       3.00       5.00      10.74      12.22
-------------------------------------------------------------------------------------------------------------

   Stats |        age     gender      unemp      urban     income    intnews   religion      remit    married
---------+---------------------------------------------------------------------------------------------------
       N |   10646.00   10646.00   10646.00   10646.00   10646.00   10646.00   10587.00    8978.00   10618.00
    Mean |      35.24       0.48       0.41       0.58       0.29       0.52       1.34       0.21       0.60
      SD |      12.32       0.50       0.49       0.49       1.84       0.50       0.66       0.41       0.49
     Min |      18.00       0.00       0.00       0.00      -5.00       0.00       1.00       0.00       0.00
     Max |      95.00       1.00       1.00       1.00      11.00       1.00       4.00       1.00       1.00
-------------------------------------------------------------------------------------------------------------

   Stats |   children   relati~s
---------+----------------------
       N |   10463.00   10608.00
    Mean |       2.21       0.37
      SD |       1.66       0.48
     Min |       0.00       0.00
     Max |       5.00       1.00
--------------------------------

. 
. *** Set 1: Economic and Political Attitudes
. * Self-Reported Economic Problems
. xtreg problem c.lnfdi_capex##c.educlvl $C1, i(ccode)

Random-effects GLS regression                   Number of obs     =     11,642
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0048                                         min =        477
     Between = 0.3026                                         avg =      831.6
     Overall = 0.0104                                         max =      1,383

                                                Wald chi2(7)      =      61.13
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0071015   .0031434     2.26   0.024     .0009406    .0132625
                educlvl |   .0012197    .005792     0.21   0.833    -.0101325    .0125719
                        |
c.lnfdi_capex#c.educlvl |  -.0030392   .0009131    -3.33   0.001    -.0048288   -.0012497
                        |
                    age |  -.0016371   .0003692    -4.43   0.000    -.0023607   -.0009136
                 gender |  -.0193202   .0094659    -2.04   0.041     -.037873   -.0007674
                  unemp |   .0129748   .0100217     1.29   0.195    -.0066674    .0326169
                  urban |   .0403956   .0099793     4.05   0.000     .0208366    .0599546
                  _cons |   .3672741   .0298712    12.30   0.000     .3087277    .4258206
------------------------+----------------------------------------------------------------
                sigma_u |  .05699746
                sigma_e |   .4622703
                    rho |  .01497498   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store problem1

. xtreg problem c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,646
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0134                                         min =        446
     Between = 0.0692                                         avg =      760.4
     Overall = 0.0145                                         max =      1,029

                                                Wald chi2(9)      =     145.18
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0059443   .0032836     1.81   0.070    -.0004915      .01238
                educlvl |   .0099307   .0060936     1.63   0.103    -.0020126     .021874
                        |
c.lnfdi_capex#c.educlvl |  -.0025876   .0009506    -2.72   0.006    -.0044507   -.0007244
                        |
                    age |  -.0015805     .00039    -4.05   0.000    -.0023449   -.0008161
                 gender |  -.0139884    .009899    -1.41   0.158      -.03339    .0054132
                  unemp |   .0252284   .0105306     2.40   0.017     .0045888     .045868
                  urban |   .0511622    .010663     4.80   0.000     .0302631    .0720613
                 income |  -.0246401   .0026849    -9.18   0.000    -.0299025   -.0193777
                intnews |    .024564    .009193     2.67   0.008      .006546    .0425819
                  _cons |   .3219699   .0345449     9.32   0.000     .2542631    .3896767
------------------------+----------------------------------------------------------------
                sigma_u |  .07879719
                sigma_e |   .4605604
                    rho |  .02843925   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store problem2

. xtreg problem c.lnfdi_capex##c.educlvl $C1 $C2 $C3, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,764
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0172                                         min =        188
     Between = 0.3396                                         avg =      626.0
     Overall = 0.0262                                         max =        907

                                                Wald chi2(14)     =     235.46
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0163807   .0035529     4.61   0.000     .0094171    .0233443
                educlvl |   .0123163   .0064031     1.92   0.054    -.0002334    .0248661
                        |
c.lnfdi_capex#c.educlvl |  -.0044148   .0010426    -4.23   0.000    -.0064583   -.0023714
                        |
                    age |   -.002001    .000462    -4.33   0.000    -.0029066   -.0010955
                 gender |  -.0085555   .0106456    -0.80   0.422    -.0294205    .0123096
                  unemp |   .0427014   .0112407     3.80   0.000     .0206699    .0647328
                  urban |   .0783819    .010616     7.38   0.000      .057575    .0991888
                 income |  -.0184498   .0027469    -6.72   0.000    -.0238336   -.0130659
                intnews |   .0353947   .0100201     3.53   0.000     .0157557    .0550338
               religion |    .009928   .0074903     1.33   0.185    -.0047528    .0246088
                  remit |  -.0176453   .0128296    -1.38   0.169    -.0427908    .0075002
                married |  -.0129997   .0113157    -1.15   0.251    -.0351781    .0091786
               children |  -.0025013   .0032675    -0.77   0.444    -.0089056    .0039029
              relatives |  -.0560982   .0109007    -5.15   0.000    -.0774633   -.0347331
                  _cons |   .2930605   .0314728     9.31   0.000      .231375     .354746
------------------------+----------------------------------------------------------------
                sigma_u |          0
                sigma_e |  .45351972
                    rho |          0   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store problem3

. xtoprobit problem c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -6781.4016  
Iteration 1:   log likelihood = -6676.1116  
Iteration 2:   log likelihood = -6676.0172  
Iteration 3:   log likelihood = -6676.0172  

Refining starting values:

Grid node 0:   log likelihood = -6580.2668

Fitting full model:

Iteration 0:   log likelihood = -6580.2668  (not concave)
Iteration 1:   log likelihood = -6575.1232  (not concave)
Iteration 2:   log likelihood = -6569.9624  (not concave)
Iteration 3:   log likelihood = -6564.8929  (not concave)
Iteration 4:   log likelihood =  -6563.958  
Iteration 5:   log likelihood = -6544.4655  
Iteration 6:   log likelihood = -6544.4549  
Iteration 7:   log likelihood = -6544.4549  

Random-effects ordered probit regression             Number of obs    = 10,646
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    446
                                                                  avg =  760.4
                                                                  max =  1,029

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     = 140.74
Log likelihood = -6544.4549                          Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0163979   .0092753     1.77   0.077    -.0017814    .0345771
                educlvl |    .031102   .0173633     1.79   0.073    -.0029295    .0651335
                        |
c.lnfdi_capex#c.educlvl |  -.0076077   .0027026    -2.81   0.005    -.0129047   -.0023108
                        |
                    age |   -.004669   .0011241    -4.15   0.000    -.0068722   -.0024658
                 gender |  -.0393389   .0284812    -1.38   0.167     -.095161    .0164832
                  unemp |   .0698041   .0301386     2.32   0.021     .0107335    .1288746
                  urban |   .1408756   .0302584     4.66   0.000     .0815702     .200181
                 income |  -.0715627   .0079155    -9.04   0.000    -.0870768   -.0560486
                intnews |   .0697817   .0263099     2.65   0.008     .0182152    .1213481
------------------------+----------------------------------------------------------------
                  /cut1 |   .4771798   .1045628                      .2722406    .6821191
------------------------+----------------------------------------------------------------
              /sigma2_u |   .0673799   .0268413                      .0308637    .1471001
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 263.12       Prob >= chibar2 = 0.0000

. est store problem4

. xtreg problem c.lnfdi_capex##c.educdev $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,828
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0128                                         min =        442
     Between = 0.1310                                         avg =      630.6
     Overall = 0.0156                                         max =        863

                                                Wald chi2(9)      =     123.10
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0015907    .001848     0.86   0.389    -.0020314    .0052129
                educdev |   .0048093   .0022321     2.15   0.031     .0004345    .0091841
                        |
c.lnfdi_capex#c.educdev |  -.0007747   .0003794    -2.04   0.041    -.0015182   -.0000311
                        |
                    age |   -.001312   .0004357    -3.01   0.003    -.0021659   -.0004581
                 gender |  -.0203891   .0106058    -1.92   0.055    -.0411761     .000398
                  unemp |   .0336191   .0112553     2.99   0.003     .0115593     .055679
                  urban |    .049304   .0114067     4.32   0.000     .0269473    .0716608
                 income |  -.0256185   .0028252    -9.07   0.000    -.0311557   -.0200813
                intnews |   .0139573   .0099998     1.40   0.163     -.005642    .0335566
                  _cons |   .3244135   .0221688    14.63   0.000     .2809634    .3678636
------------------------+----------------------------------------------------------------
                sigma_u |  .02708672
                sigma_e |   .4566251
                    rho |  .00350645   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store problem5

. xtreg problem c.lnfdi_jobs##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,646
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0135                                         min =        446
     Between = 0.0535                                         avg =      760.4
     Overall = 0.0141                                         max =      1,029

                                                Wald chi2(9)      =     145.87
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0048993   .0028186     1.74   0.082     -.000625    .0104235
               educlvl |   .0108239   .0063256     1.71   0.087    -.0015741    .0232218
                       |
c.lnfdi_jobs#c.educlvl |  -.0023069   .0008344    -2.76   0.006    -.0039423   -.0006715
                       |
                   age |  -.0015775   .0003899    -4.05   0.000    -.0023418   -.0008132
                gender |  -.0141009   .0098976    -1.42   0.154    -.0334999    .0052981
                 unemp |   .0256109   .0105279     2.43   0.015     .0049765    .0462453
                 urban |   .0515841   .0106369     4.85   0.000      .030736    .0724321
                income |  -.0245368   .0026856    -9.14   0.000    -.0298004   -.0192731
               intnews |   .0244042   .0091891     2.66   0.008      .006394    .0424145
                 _cons |   .3219175   .0344186     9.35   0.000     .2544583    .3893768
-----------------------+----------------------------------------------------------------
               sigma_u |  .07542924
               sigma_e |  .46054037
                   rho |  .02612447   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store problem6

. * Expected Social Decline
. xtreg worse c.lnfdi_capex##c.educlvl $C1, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,425
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0064                                         min =        386
     Between = 0.1291                                         avg =      744.6
     Overall = 0.0092                                         max =      1,137

                                                Wald chi2(7)      =      68.10
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                  worse | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |    .007681   .0140641     0.55   0.585    -.0198841    .0352461
                educlvl |     .05258   .0260007     2.02   0.043     .0016195    .1035406
                        |
c.lnfdi_capex#c.educlvl |   -.006844   .0040669    -1.68   0.092     -.014815     .001127
                        |
                    age |   .0116825   .0016516     7.07   0.000     .0084455    .0149196
                 gender |   .1391699     .04186     3.32   0.001     .0571258     .221214
                  unemp |  -.0433911   .0444089    -0.98   0.329    -.1304311    .0436488
                  urban |  -.0304198   .0447894    -0.68   0.497    -.1182055    .0573658
                  _cons |   -1.91156   .1552264   -12.31   0.000    -2.215798   -1.607321
------------------------+----------------------------------------------------------------
                sigma_u |  .38710899
                sigma_e |   1.936942
                    rho |  .03840821   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store worse1

. xtreg worse c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      9,661
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0064                                         min =        365
     Between = 0.1297                                         avg =      690.1
     Overall = 0.0090                                         max =        886

                                                Wald chi2(9)      =      62.86
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                  worse | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0132103   .0146175     0.90   0.366    -.0154394      .04186
                educlvl |    .056551   .0271405     2.08   0.037     .0033566    .1097454
                        |
c.lnfdi_capex#c.educlvl |  -.0082276   .0042072    -1.96   0.051    -.0164736    .0000183
                        |
                    age |   .0110923    .001734     6.40   0.000     .0076937    .0144908
                 gender |   .1288678   .0437296     2.95   0.003     .0431593    .2145763
                  unemp |  -.0470409   .0464929    -1.01   0.312    -.1381653    .0440835
                  urban |  -.0517529   .0475957    -1.09   0.277    -.1450387    .0415329
                 income |   .0112401   .0117153     0.96   0.337    -.0117216    .0342017
                intnews |  -.0963971   .0406422    -2.37   0.018    -.1760543     -.01674
                  _cons |  -1.833556   .1750198   -10.48   0.000    -2.176589   -1.490524
------------------------+----------------------------------------------------------------
                sigma_u |  .46837985
                sigma_e |   1.937405
                    rho |  .05521878   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store worse2

. xtreg worse c.lnfdi_capex##c.educlvl $C1 $C2 $C3, i(ccode)

Random-effects GLS regression                   Number of obs     =      7,989
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0038                                         min =        159
     Between = 0.3970                                         avg =      570.6
     Overall = 0.0276                                         max =        827

                                                Wald chi2(14)     =     226.75
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                  worse | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0581718   .0165813     3.51   0.000      .025673    .0906705
                educlvl |  -.0040958    .030208    -0.14   0.892    -.0633025    .0551109
                        |
c.lnfdi_capex#c.educlvl |  -.0123374   .0048664    -2.54   0.011    -.0218753   -.0027994
                        |
                    age |   .0122262   .0021583     5.66   0.000      .007996    .0164564
                 gender |   .1988956   .0495085     4.02   0.000     .1018606    .2959305
                  unemp |  -.1042065   .0522827    -1.99   0.046    -.2066787   -.0017344
                  urban |   .0898096   .0494899     1.81   0.070    -.0071888    .1868081
                 income |   .0285921   .0126139     2.27   0.023     .0038693    .0533149
                intnews |  -.0926821   .0465361    -1.99   0.046    -.1838911   -.0014731
               religion |   .1457654   .0344628     4.23   0.000     .0782196    .2133111
                  remit |   .0307887   .0590482     0.52   0.602    -.0849437     .146521
                married |   .2654762   .0526534     5.04   0.000     .1622773     .368675
               children |  -.0498159   .0153051    -3.25   0.001    -.0798134   -.0198183
              relatives |  -.1358163   .0506139    -2.68   0.007    -.2350177   -.0366149
                  _cons |  -2.092716   .1479791   -14.14   0.000     -2.38275   -1.802683
------------------------+----------------------------------------------------------------
                sigma_u |          0
                sigma_e |  1.9220126
                    rho |          0   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store worse3

. xtoprobit worse c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -20319.565  
Iteration 1:   log likelihood = -20219.781  
Iteration 2:   log likelihood =  -20219.78  

Refining starting values:

Grid node 0:   log likelihood = -19744.686

Fitting full model:

Iteration 0:   log likelihood = -19744.686  (not concave)
Iteration 1:   log likelihood = -19741.319  (not concave)
Iteration 2:   log likelihood = -19737.951  (not concave)
Iteration 3:   log likelihood = -19735.049  (not concave)
Iteration 4:   log likelihood = -19732.606  (not concave)
Iteration 5:   log likelihood = -19731.662  
Iteration 6:   log likelihood = -19657.685  
Iteration 7:   log likelihood = -19657.178  
Iteration 8:   log likelihood = -19657.148  
Iteration 9:   log likelihood = -19657.148  

Random-effects ordered probit regression             Number of obs    =  9,661
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    365
                                                                  avg =  690.1
                                                                  max =    886

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     =  67.46
Log likelihood = -19657.148                          Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
                  worse | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0049481   .0076593     0.65   0.518    -.0100638    .0199601
                educlvl |   .0246874   .0142108     1.74   0.082    -.0031653    .0525401
                        |
c.lnfdi_capex#c.educlvl |  -.0037836   .0022019    -1.72   0.086    -.0080993    .0005321
                        |
                    age |   .0061385   .0009082     6.76   0.000     .0043584    .0079185
                 gender |   .0652061   .0228768     2.85   0.004     .0203684    .1100438
                  unemp |  -.0164135   .0243243    -0.67   0.500    -.0640883    .0312612
                  urban |  -.0325135   .0249567    -1.30   0.193    -.0814278    .0164008
                 income |   .0066842   .0061378     1.09   0.276    -.0053457    .0187141
                intnews |  -.0472753   .0212614    -2.22   0.026    -.0889469   -.0056037
------------------------+----------------------------------------------------------------
                  /cut1 |   -2.97217   .1493548                       -3.2649   -2.679441
                  /cut2 |  -2.699437   .1353418                     -2.964702   -2.434172
                  /cut3 |  -2.521214   .1303295                     -2.776655   -2.265773
                  /cut4 |  -2.249708   .1259037                     -2.496474   -2.002941
                  /cut5 |  -1.936282   .1232747                     -2.177896   -1.694668
                  /cut6 |  -1.440203   .1214179                     -1.678177   -1.202228
                  /cut7 |   -1.00035   .1207542                     -1.237024   -.7636764
                  /cut8 |  -.5058591   .1204421                     -.7419213   -.2697969
                  /cut9 |   .1870194   .1203366                      -.048836    .4228748
                 /cut10 |   .7938748   .1204801                      .5577382    1.030011
                 /cut11 |   1.433667   .1209507                      1.196607    1.670726
                 /cut12 |   1.796545   .1214707                      1.558466    2.034623
                 /cut13 |   2.214337   .1225872                       1.97407    2.454603
                 /cut14 |   2.461712   .1238263                      2.219017    2.704407
                 /cut15 |    2.74525   .1265024                       2.49731     2.99319
                 /cut16 |   3.020688    .131728                      2.762506    3.278871
                 /cut17 |   3.197103   .1377028                      2.927211    3.466996
                 /cut18 |   3.436558   .1519836                      3.138676     3.73444
                 /cut19 |    3.63209   .1726988                      3.293607    3.970574
                 /cut20 |   3.878723   .2206095                      3.446336    4.311109
------------------------+----------------------------------------------------------------
              /sigma2_u |   .1445024   .0553098                      .0682447    .3059719
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 1125.26      Prob >= chibar2 = 0.0000

. est store worse4

. xtreg worse c.lnfdi_capex##c.educdev $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,111
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0072                                         min =        362
     Between = 0.0039                                         avg =      579.4
     Overall = 0.0066                                         max =        815

                                                Wald chi2(9)      =      59.09
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                  worse | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |  -.0191717   .0090581    -2.12   0.034    -.0369253   -.0014182
                educdev |   .0074303   .0098682     0.75   0.451    -.0119111    .0267716
                        |
c.lnfdi_capex#c.educdev |   .0005706   .0016807     0.34   0.734    -.0027236    .0038647
                        |
                    age |   .0123282   .0019556     6.30   0.000     .0084953    .0161611
                 gender |   .1355284   .0473036     2.87   0.004      .042815    .2282418
                  unemp |  -.0149534   .0505467    -0.30   0.767    -.1140231    .0841163
                  urban |  -.0689133   .0534699    -1.29   0.197    -.1737125    .0358858
                 income |   .0120328   .0126639     0.95   0.342     -.012788    .0368535
                intnews |  -.0755249   .0445139    -1.70   0.090    -.1627706    .0117208
                  _cons |  -1.694503   .2496785    -6.79   0.000    -2.183864   -1.205142
------------------------+----------------------------------------------------------------
                sigma_u |  .86096376
                sigma_e |  1.9513032
                    rho |  .16295542   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store worse5

. xtreg worse c.lnfdi_jobs##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      9,661
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0065                                         min =        365
     Between = 0.0747                                         avg =      690.1
     Overall = 0.0084                                         max =        886

                                                Wald chi2(9)      =      63.35
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
                 worse | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0033231   .0125409     0.26   0.791    -.0212566    .0279029
               educlvl |   .0509927   .0281465     1.81   0.070    -.0041734    .1061588
                       |
c.lnfdi_jobs#c.educlvl |  -.0059374   .0036894    -1.61   0.108    -.0131685    .0012936
                       |
                   age |   .0110675   .0017335     6.38   0.000     .0076699    .0144651
                gender |   .1274109   .0437155     2.91   0.004     .0417301    .2130917
                 unemp |  -.0451669   .0464816    -0.97   0.331    -.1362692    .0459354
                 urban |  -.0502717   .0475433    -1.06   0.290    -.1434549    .0429115
                income |   .0116395   .0117204     0.99   0.321    -.0113322    .0346112
               intnews |  -.0974771   .0406156    -2.40   0.016    -.1770821   -.0178721
                 _cons |  -1.787446   .1881857    -9.50   0.000    -2.156283   -1.418608
-----------------------+----------------------------------------------------------------
               sigma_u |  .52775433
               sigma_e |  1.9373389
                   rho |  .06908184   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store worse6

. * Leader Satisfaction
. xtreg leadsat c.lnfdi_capex##c.educlvl $C1, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,980
Group variable: ccode                           Number of groups  =         13

R-squared:                                      Obs per group:
     Within  = 0.0079                                         min =        602
     Between = 0.0841                                         avg =      844.6
     Overall = 0.0004                                         max =      1,323

                                                Wald chi2(7)      =      84.62
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                leadsat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |  -.0161005   .0062568    -2.57   0.010    -.0283635   -.0038374
                educlvl |   .0224205   .0116271     1.93   0.054    -.0003682    .0452091
                        |
c.lnfdi_capex#c.educlvl |  -.0039683   .0018316    -2.17   0.030    -.0075581   -.0003785
                        |
                    age |  -.0012596    .000737    -1.71   0.087    -.0027042    .0001849
                 gender |   .0087377   .0189456     0.46   0.645    -.0283949    .0458704
                  unemp |    .016203   .0200254     0.81   0.418     -.023046    .0554521
                  urban |   .0950949   .0198067     4.80   0.000     .0562744    .1339154
                  _cons |   2.028978   .1029874    19.70   0.000     1.827127     2.23083
------------------------+----------------------------------------------------------------
                sigma_u |  .32135122
                sigma_e |  .89852676
                    rho |   .1134029   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store leadsat1

. xtreg leadsat c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,058
Group variable: ccode                           Number of groups  =         13

R-squared:                                      Obs per group:
     Within  = 0.0063                                         min =        568
     Between = 0.0527                                         avg =      773.7
     Overall = 0.0000                                         max =        982

                                                Wald chi2(9)      =      61.64
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                leadsat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |  -.0086332   .0064688    -1.33   0.182    -.0213118    .0040454
                educlvl |   .0301025    .012136     2.48   0.013     .0063165    .0538886
                        |
c.lnfdi_capex#c.educlvl |  -.0041362     .00189    -2.19   0.029    -.0078406   -.0004318
                        |
                    age |  -.0010496   .0007723    -1.36   0.174    -.0025633    .0004641
                 gender |   .0117497   .0196711     0.60   0.550     -.026805    .0503045
                  unemp |   .0041708   .0208789     0.20   0.842    -.0367511    .0450927
                  urban |   .0889107   .0209513     4.24   0.000      .047847    .1299744
                 income |   -.011629   .0052812    -2.20   0.028    -.0219799    -.001278
                intnews |    .029351   .0183438     1.60   0.110    -.0066023    .0653042
                  _cons |   1.945332   .1070942    18.16   0.000     1.735431    2.155233
------------------------+----------------------------------------------------------------
                sigma_u |  .33226406
                sigma_e |  .89008209
                    rho |  .12230649   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store leadsat2

. xtreg leadsat c.lnfdi_capex##c.educlvl $C1 $C2 $C3, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,229
Group variable: ccode                           Number of groups  =         13

R-squared:                                      Obs per group:
     Within  = 0.0004                                         min =        178
     Between = 0.5214                                         avg =      633.0
     Overall = 0.0137                                         max =        885

                                                Wald chi2(14)     =     114.03
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                leadsat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0278125   .0077283     3.60   0.000     .0126653    .0429598
                educlvl |   .0748674   .0144194     5.19   0.000     .0466059     .103129
                        |
c.lnfdi_capex#c.educlvl |  -.0061034   .0023037    -2.65   0.008    -.0106186   -.0015883
                        |
                    age |  -.0007204   .0010131    -0.71   0.477     -.002706    .0012652
                 gender |   .0830716    .023524     3.53   0.000     .0369654    .1291778
                  unemp |   -.072888   .0247223    -2.95   0.003    -.1213428   -.0244333
                  urban |  -.0184858   .0239962    -0.77   0.441    -.0655174    .0285459
                 income |  -.0095366   .0059829    -1.59   0.111    -.0212628    .0021896
                intnews |   .0587929   .0222228     2.65   0.008      .015237    .1023487
               religion |  -.0001264   .0162549    -0.01   0.994    -.0319854    .0317327
                  remit |   .1179869   .0282047     4.18   0.000     .0627067    .1732671
                married |  -.0879667   .0249971    -3.52   0.000    -.1369601   -.0389734
               children |   .0282188   .0072426     3.90   0.000     .0140236     .042414
              relatives |  -.0574748   .0244297    -2.35   0.019    -.1053561   -.0095935
                  _cons |   1.690075   .0692719    24.40   0.000     1.554305    1.825846
------------------------+----------------------------------------------------------------
                sigma_u |          0
                sigma_e |  .88320073
                    rho |          0   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store leadsat3

. xtoprobit leadsat c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -12681.668  
Iteration 1:   log likelihood = -12611.512  
Iteration 2:   log likelihood = -12611.495  
Iteration 3:   log likelihood = -12611.495  

Refining starting values:

Grid node 0:   log likelihood = -11718.156

Fitting full model:

Iteration 0:   log likelihood = -11718.156  (not concave)
Iteration 1:   log likelihood = -11714.383  (not concave)
Iteration 2:   log likelihood = -11710.623  (not concave)
Iteration 3:   log likelihood =  -11709.88  
Iteration 4:   log likelihood = -11614.413  
Iteration 5:   log likelihood =  -11578.68  
Iteration 6:   log likelihood = -11578.387  
Iteration 7:   log likelihood = -11578.376  
Iteration 8:   log likelihood = -11578.376  

Random-effects ordered probit regression             Number of obs    = 10,058
Group variable: ccode                                Number of groups =     13

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    568
                                                                  avg =  773.7
                                                                  max =    982

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     =  53.21
Log likelihood = -11578.376                          Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
                leadsat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |  -.0063723   .0083766    -0.76   0.447    -.0227901    .0100456
                educlvl |     .04037   .0156299     2.58   0.010     .0097359     .071004
                        |
c.lnfdi_capex#c.educlvl |  -.0053062   .0024232    -2.19   0.029    -.0100556   -.0005568
                        |
                    age |  -.0011824   .0009777    -1.21   0.226    -.0030986    .0007338
                 gender |   .0194437   .0247081     0.79   0.431    -.0289832    .0678706
                  unemp |   .0125304   .0262516     0.48   0.633    -.0389219    .0639826
                  urban |   .1205257   .0265122     4.55   0.000     .0685627    .1724887
                 income |  -.0138091   .0066044    -2.09   0.037    -.0267534   -.0008648
                intnews |   .0271915    .023112     1.18   0.239    -.0181072    .0724903
------------------------+----------------------------------------------------------------
                  /cut1 |  -.2291822   .1730844                     -.5684214    .1100571
                  /cut2 |   .8278398   .1732208                      .4883333    1.167346
                  /cut3 |   1.468959   .1735063                      1.128893    1.809025
------------------------+----------------------------------------------------------------
              /sigma2_u |   .3258566   .1287053                      .1502529    .7066917
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 2066.24      Prob >= chibar2 = 0.0000

. est store leadsat4

. xtreg leadsat c.lnfdi_capex##c.educdev $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,299
Group variable: ccode                           Number of groups  =         13

R-squared:                                      Obs per group:
     Within  = 0.0113                                         min =        494
     Between = 0.1032                                         avg =      638.4
     Overall = 0.0001                                         max =        864

                                                Wald chi2(9)      =      93.77
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                leadsat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |    -.03147   .0040466    -7.78   0.000    -.0394012   -.0235387
                educdev |    .007893   .0044032     1.79   0.073    -.0007371    .0165231
                        |
c.lnfdi_capex#c.educdev |  -.0017129   .0007429    -2.31   0.021    -.0031689   -.0002569
                        |
                    age |  -.0008679    .000856    -1.01   0.311    -.0025457    .0008098
                 gender |   .0177359    .020931     0.85   0.397    -.0232881    .0587599
                  unemp |  -.0312661   .0223342    -1.40   0.162    -.0750403    .0125081
                  urban |   .1119555   .0230553     4.86   0.000      .066768    .1571431
                 income |  -.0065156    .005565    -1.17   0.242    -.0174228    .0043917
                intnews |   .0505235   .0197474     2.56   0.011     .0118193    .0892277
                  _cons |   2.078613   .1522032    13.66   0.000     1.780301    2.376926
------------------------+----------------------------------------------------------------
                sigma_u |  .52595054
                sigma_e |  .87387208
                    rho |  .26591392   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store leadsat5

. xtreg leadsat c.lnfdi_jobs##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,058
Group variable: ccode                           Number of groups  =         13

R-squared:                                      Obs per group:
     Within  = 0.0049                                         min =        568
     Between = 0.0354                                         avg =      773.7
     Overall = 0.0000                                         max =        982

                                                Wald chi2(9)      =      48.61
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
               leadsat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |  -.0075129   .0055477    -1.35   0.176    -.0183862    .0033604
               educlvl |    .023244   .0126056     1.84   0.065    -.0014626    .0479506
                       |
c.lnfdi_jobs#c.educlvl |  -.0024612   .0016615    -1.48   0.139    -.0057176    .0007951
                       |
                   age |  -.0010539   .0007727    -1.36   0.173    -.0025683    .0004605
                gender |   .0107446   .0196786     0.55   0.585    -.0278248    .0493139
                 unemp |   .0052984   .0208879     0.25   0.800    -.0356411    .0462378
                 urban |   .0848877   .0209326     4.06   0.000     .0438606    .1259149
                income |  -.0118949   .0052866    -2.25   0.024    -.0222564   -.0015333
               intnews |   .0316455   .0183459     1.72   0.085    -.0043119    .0676028
                 _cons |   1.948684   .1142405    17.06   0.000     1.724777    2.172592
-----------------------+----------------------------------------------------------------
               sigma_u |  .35999429
               sigma_e |  .89068648
                   rho |  .14041968   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store leadsat6

. 
. *** TABLE A-14: Economic and Political Attitudes - Main Models
. esttab problem2 problem3 problem4 worse2 worse3 worse4 leadsat2 leadsat3 leadsat4, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl lnfdi_capex c.lnfdi_capex#c.educlvl age gender unemp urban income intnews ///
>                 religion remit married children relatives) ///
>         drop(_cons) eqlabels("" "") nogaps

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                  problem         problem         problem           worse           worse           worse         le
> adsat         leadsat         leadsat   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
educlvl             0.010           0.012*          0.031*          0.057**        -0.004           0.025*          
> 0.030**         0.075***        0.040***
                   (0.01)          (0.01)          (0.02)          (0.03)          (0.03)          (0.01)          (
> 0.01)          (0.01)          (0.02)   
lnfdi_capex         0.006*          0.016***        0.016*          0.013           0.058***        0.005          -
> 0.009           0.028***       -0.006   
                   (0.00)          (0.00)          (0.01)          (0.01)          (0.02)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
c.lnfdi_ca~l       -0.003***       -0.004***       -0.008***       -0.008*         -0.012**        -0.004*         -
> 0.004**        -0.006***       -0.005** 
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
age                -0.002***       -0.002***       -0.005***        0.011***        0.012***        0.006***       -
> 0.001          -0.001          -0.001   
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
gender             -0.014          -0.009          -0.039           0.129***        0.199***        0.065***        
> 0.012           0.083***        0.019   
                   (0.01)          (0.01)          (0.03)          (0.04)          (0.05)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
unemp               0.025**         0.043***        0.070**        -0.047          -0.104**        -0.016           
> 0.004          -0.073***        0.013   
                   (0.01)          (0.01)          (0.03)          (0.05)          (0.05)          (0.02)          (
> 0.02)          (0.02)          (0.03)   
urban               0.051***        0.078***        0.141***       -0.052           0.090*         -0.033           
> 0.089***       -0.018           0.121***
                   (0.01)          (0.01)          (0.03)          (0.05)          (0.05)          (0.02)          (
> 0.02)          (0.02)          (0.03)   
income             -0.025***       -0.018***       -0.072***        0.011           0.029**         0.007          -
> 0.012**        -0.010          -0.014** 
                   (0.00)          (0.00)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
intnews             0.025***        0.035***        0.070***       -0.096**        -0.093**        -0.047**         
> 0.029           0.059***        0.027   
                   (0.01)          (0.01)          (0.03)          (0.04)          (0.05)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
religion                            0.010                                           0.146***                        
>                -0.000                   
                                   (0.01)                                          (0.03)                           
>                (0.02)                   
remit                              -0.018                                           0.031                           
>                 0.118***                
                                   (0.01)                                          (0.06)                           
>                (0.03)                   
married                            -0.013                                           0.265***                        
>                -0.088***                
                                   (0.01)                                          (0.05)                           
>                (0.02)                   
children                           -0.003                                          -0.050***                        
>                 0.028***                
                                   (0.00)                                          (0.02)                           
>                (0.01)                   
relatives                          -0.056***                                       -0.136***                        
>                -0.057**                 
                                   (0.01)                                          (0.05)                           
>                (0.02)                   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        10646            8764           10646            9661            7989            9661           
> 10058            8229           10058   
\# of coun~s           14              14              14              14              14              14           
>    13              13              13   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

.         
. *** TABLE A-15: Economic and Political Attitudes - Robustness Tests
. esttab problem1 problem5 problem6 worse1 worse5 worse6 leadsat1 leadsat5 leadsat6, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl educdev lnfdi_capex c.lnfdi_capex#c.educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl c.lnfdi_capex#
> c.educdev ///
>                 age gender unemp urban income intnews) ///
>         drop(_cons) eqlabels("" "") nogaps

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
                  problem         problem         problem           worse           worse           worse         le
> adsat         leadsat         leadsat   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
educlvl             0.001                           0.011*          0.053**                         0.051*          
> 0.022*                          0.023*  
                   (0.01)                          (0.01)          (0.03)                          (0.03)          (
> 0.01)                          (0.01)   
educdev                             0.005**                                         0.007                           
>                 0.008*                  
                                   (0.00)                                          (0.01)                           
>                (0.00)                   
lnfdi_capex         0.007**         0.002                           0.008          -0.019**                        -
> 0.016**        -0.031***                
                   (0.00)          (0.00)                          (0.01)          (0.01)                          (
> 0.01)          (0.00)                   
c.lnfdi_ca~l       -0.003***                                       -0.007*                                         -
> 0.004**                                 
                   (0.00)                                          (0.00)                                          (
> 0.00)                                   
lnfdi_jobs                                          0.005*                                          0.003           
>                                -0.008   
                                                   (0.00)                                          (0.01)           
>                                (0.01)   
c.lnfdi_jo~l                                       -0.002***                                       -0.006           
>                                -0.002   
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
c.lnfdi_ca~v                       -0.001**                                         0.001                           
>                -0.002**                 
                                   (0.00)                                          (0.00)                           
>                (0.00)                   
age                -0.002***       -0.001***       -0.002***        0.012***        0.012***        0.011***       -
> 0.001*         -0.001          -0.001   
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
gender             -0.019**        -0.020*         -0.014           0.139***        0.136***        0.127***        
> 0.009           0.018           0.011   
                   (0.01)          (0.01)          (0.01)          (0.04)          (0.05)          (0.04)          (
> 0.02)          (0.02)          (0.02)   
unemp               0.013           0.034***        0.026**        -0.043          -0.015          -0.045           
> 0.016          -0.031           0.005   
                   (0.01)          (0.01)          (0.01)          (0.04)          (0.05)          (0.05)          (
> 0.02)          (0.02)          (0.02)   
urban               0.040***        0.049***        0.052***       -0.030          -0.069          -0.050           
> 0.095***        0.112***        0.085***
                   (0.01)          (0.01)          (0.01)          (0.04)          (0.05)          (0.05)          (
> 0.02)          (0.02)          (0.02)   
income                             -0.026***       -0.025***                        0.012           0.012           
>                -0.007          -0.012** 
                                   (0.00)          (0.00)                          (0.01)          (0.01)           
>                (0.01)          (0.01)   
intnews                             0.014           0.024***                       -0.076*         -0.097**         
>                 0.051**         0.032*  
                                   (0.01)          (0.01)                          (0.04)          (0.04)           
>                (0.02)          (0.02)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        11642            8828           10646           10425            8111            9661           
> 10980            8299           10058   
\# of coun~s           14              14              14              14              14              14           
>    13              13              13   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** Set 2: Perceived Regime Legitimacy
. * Legitimacy Belief: Freedom of Speech
. xtreg legit_free c.lnfdi_capex##c.educlvl $C1, i(ccode)

Random-effects GLS regression                   Number of obs     =     11,442
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0038                                         min =        441
     Between = 0.0374                                         avg =      817.3
     Overall = 0.0060                                         max =      1,278

                                                Wald chi2(7)      =      43.35
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0380918   .0064853     5.87   0.000     .0253809    .0508028
                educlvl |   .0518407   .0118442     4.38   0.000     .0286265    .0750549
                        |
c.lnfdi_capex#c.educlvl |  -.0095024    .001874    -5.07   0.000    -.0131754   -.0058294
                        |
                    age |  -.0010764   .0007546    -1.43   0.154    -.0025554    .0004026
                 gender |   .0017469   .0193087     0.09   0.928    -.0360975    .0395913
                  unemp |   .0240849   .0204599     1.18   0.239    -.0160158    .0641857
                  urban |   .0296278   .0205951     1.44   0.150    -.0107379    .0699935
                  _cons |   .6364567   .1215653     5.24   0.000      .398193    .8747203
------------------------+----------------------------------------------------------------
                sigma_u |  .40976669
                sigma_e |  .93587749
                    rho |  .16086673   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_free1

. xtreg legit_free c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,480
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0044                                         min =        416
     Between = 0.0513                                         avg =      748.6
     Overall = 0.0072                                         max =        946

                                                Wald chi2(9)      =      46.35
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0403298   .0067375     5.99   0.000     .0271246     .053535
                educlvl |   .0589642   .0124389     4.74   0.000     .0345843     .083344
                        |
c.lnfdi_capex#c.educlvl |  -.0108482   .0019428    -5.58   0.000    -.0146561   -.0070404
                        |
                    age |  -.0012338   .0007944    -1.55   0.120    -.0027909    .0003232
                 gender |   .0041273    .020122     0.21   0.837    -.0353111    .0435658
                  unemp |   .0204996   .0214193     0.96   0.339    -.0214815    .0624806
                  urban |   .0297488   .0218841     1.36   0.174    -.0131432    .0726409
                 income |   .0035284   .0054681     0.65   0.519    -.0071888    .0142457
                intnews |  -.0012428   .0187543    -0.07   0.947    -.0380005    .0355149
                  _cons |    .635365   .1043603     6.09   0.000     .4308225    .8399075
------------------------+----------------------------------------------------------------
                sigma_u |   .3293814
                sigma_e |  .93057463
                    rho |  .11133547   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_free2

. xtreg legit_free c.lnfdi_capex##c.educlvl $C1 $C2 $C3, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,686
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0037                                         min =        179
     Between = 0.2914                                         avg =      620.4
     Overall = 0.0191                                         max =        895

                                                Wald chi2(14)     =     168.77
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0619769    .007572     8.18   0.000      .047136    .0768178
                educlvl |   .0275301   .0136567     2.02   0.044     .0007634    .0542968
                        |
c.lnfdi_capex#c.educlvl |  -.0129737   .0022196    -5.85   0.000    -.0173239   -.0086234
                        |
                    age |  -.0038537   .0009856    -3.91   0.000    -.0057854   -.0019219
                 gender |   .0122848   .0226816     0.54   0.588    -.0321704      .05674
                  unemp |   .0368822   .0239497     1.54   0.124    -.0100584    .0838228
                  urban |   .0003384   .0226695     0.01   0.988     -.044093    .0447698
                 income |   .0242018   .0058488     4.14   0.000     .0127384    .0356651
                intnews |    .070506   .0213823     3.30   0.001     .0285974    .1124147
               religion |  -.0389004   .0159287    -2.44   0.015      -.07012   -.0076808
                  remit |   .0406818   .0273587     1.49   0.137    -.0129403    .0943039
                married |  -.0086492   .0241489    -0.36   0.720    -.0559803    .0386818
               children |   .0158673   .0069756     2.27   0.023     .0021953    .0295393
              relatives |   .0706858   .0232793     3.04   0.002     .0250592    .1163124
                  _cons |   .7360933   .0671287    10.97   0.000     .6045234    .8676632
------------------------+----------------------------------------------------------------
                sigma_u |          0
                sigma_e |  .91886083
                    rho |          0   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_free3

. xtoprobit legit_free c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -12752.828  
Iteration 1:   log likelihood = -12691.103  
Iteration 2:   log likelihood = -12691.096  
Iteration 3:   log likelihood = -12691.096  

Refining starting values:

Grid node 0:   log likelihood = -12216.297

Fitting full model:

Iteration 0:   log likelihood = -12216.297  (not concave)
Iteration 1:   log likelihood = -12211.824  (not concave)
Iteration 2:   log likelihood = -12209.787  (not concave)
Iteration 3:   log likelihood = -12208.109  
Iteration 4:   log likelihood = -12199.841  
Iteration 5:   log likelihood = -12175.222  
Iteration 6:   log likelihood =   -12175.2  
Iteration 7:   log likelihood =   -12175.2  

Random-effects ordered probit regression             Number of obs    = 10,480
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    416
                                                                  avg =  748.6
                                                                  max =    946

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     =  49.37
Log likelihood = -12175.2                            Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
             legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0512057   .0082324     6.22   0.000     .0350704    .0673409
                educlvl |   .0770005   .0152234     5.06   0.000     .0471632    .1068379
                        |
c.lnfdi_capex#c.educlvl |  -.0138897   .0023752    -5.85   0.000     -.018545   -.0092345
                        |
                    age |  -.0013382   .0009695    -1.38   0.168    -.0032384    .0005621
                 gender |   .0063591   .0244672     0.26   0.795    -.0415957     .054314
                  unemp |   .0198601   .0259762     0.76   0.445    -.0310524    .0707726
                  urban |   .0375057   .0264462     1.42   0.156    -.0143279    .0893393
                 income |   .0058136    .006595     0.88   0.378    -.0071123    .0187395
                intnews |   .0069949   .0228334     0.31   0.759    -.0377576    .0517475
------------------------+----------------------------------------------------------------
                  /cut1 |   .2453999   .1285916                      -.006635    .4974348
                  /cut2 |   1.020493   .1288377                      .7679755     1.27301
                  /cut3 |   1.764555   .1294513                      1.510835    2.018274
------------------------+----------------------------------------------------------------
              /sigma2_u |   .1652248   .0633916                      .0778925    .3504732
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 1031.79      Prob >= chibar2 = 0.0000

. est store legit_free4

. xtreg legit_free c.lnfdi_capex##c.educdev $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,752
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0039                                         min =        412
     Between = 0.1534                                         avg =      625.1
     Overall = 0.0073                                         max =        859

                                                Wald chi2(9)      =      35.17
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0001

-----------------------------------------------------------------------------------------
             legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0030072   .0041142     0.73   0.465    -.0050565     .011071
                educdev |    .014608   .0045659     3.20   0.001     .0056589     .023557
                        |
c.lnfdi_capex#c.educdev |  -.0034365    .000774    -4.44   0.000    -.0049534   -.0019195
                        |
                    age |  -.0014868   .0008876    -1.68   0.094    -.0032265    .0002529
                 gender |   .0256117   .0215944     1.19   0.236    -.0167126     .067936
                  unemp |   .0075704   .0231099     0.33   0.743    -.0377243    .0528651
                  urban |   .0450573    .024305     1.85   0.064    -.0025796    .0926942
                 income |   .0112683   .0058247     1.93   0.053    -.0001479    .0226845
                intnews |   .0163851   .0203611     0.80   0.421    -.0235219    .0562921
                  _cons |   .8148382   .0925878     8.80   0.000     .6333695    .9963068
------------------------+----------------------------------------------------------------
                sigma_u |  .30474597
                sigma_e |  .92548998
                    rho |   .0978196   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_free5

. xtreg legit_free c.lnfdi_jobs##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,480
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0040                                         min =        416
     Between = 0.0680                                         avg =      748.6
     Overall = 0.0075                                         max =        946

                                                Wald chi2(9)      =      42.24
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
            legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0328162   .0057751     5.68   0.000     .0214971    .0441353
               educlvl |   .0575199   .0128993     4.46   0.000     .0322378     .082802
                       |
c.lnfdi_jobs#c.educlvl |   -.008754   .0017042    -5.14   0.000    -.0120941   -.0054139
                       |
                   age |  -.0011994   .0007946    -1.51   0.131    -.0027567     .000358
                gender |   .0037438   .0201265     0.19   0.852    -.0357035     .043191
                 unemp |   .0213079   .0214217     0.99   0.320    -.0206777    .0632936
                 urban |   .0308326   .0218455     1.41   0.158    -.0119837    .0736489
                income |   .0035639   .0054721     0.65   0.515    -.0071612    .0142889
               intnews |  -.0014027   .0187532    -0.07   0.940    -.0381584     .035353
                 _cons |   .6363721   .0984631     6.46   0.000     .4433881    .8293561
-----------------------+----------------------------------------------------------------
               sigma_u |     .30018
               sigma_e |  .93076333
                   rho |  .09421304   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store legit_free6

. * Legitimacy Belief: Judiciary
. xtreg legit_judi c.lnfdi_capex##c.educlvl $C1, i(ccode)

Random-effects GLS regression                   Number of obs     =     11,365
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0058                                         min =        444
     Between = 0.0176                                         avg =      811.8
     Overall = 0.0062                                         max =      1,306

                                                Wald chi2(7)      =      66.54
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0168679   .0074091     2.28   0.023     .0023463    .0313895
                educlvl |   .0683962   .0136342     5.02   0.000     .0416736    .0951187
                        |
c.lnfdi_capex#c.educlvl |   -.007333   .0021421    -3.42   0.001    -.0115314   -.0031346
                        |
                    age |   .0023855   .0008595     2.78   0.006     .0007008    .0040702
                 gender |   .0140329   .0219843     0.64   0.523    -.0290555    .0571213
                  unemp |  -.0311439    .023318    -1.34   0.182    -.0768462    .0145585
                  urban |   .1193122   .0234285     5.09   0.000     .0733932    .1652312
                  _cons |   .8092271   .1182861     6.84   0.000     .5773907    1.041064
------------------------+----------------------------------------------------------------
                sigma_u |  .38013513
                sigma_e |  1.0615664
                    rho |  .11365405   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_judi1

. xtreg legit_judi c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,407
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0080                                         min =        415
     Between = 0.0274                                         avg =      743.4
     Overall = 0.0089                                         max =        973

                                                Wald chi2(9)      =      83.57
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0137792   .0077234     1.78   0.074    -.0013584    .0289168
                educlvl |   .0681143   .0143506     4.75   0.000     .0399877    .0962409
                        |
c.lnfdi_capex#c.educlvl |  -.0081975   .0022278    -3.68   0.000    -.0125639   -.0038311
                        |
                    age |   .0025428   .0009064     2.81   0.005     .0007663    .0043193
                 gender |   .0143746   .0229721     0.63   0.531    -.0306498     .059399
                  unemp |  -.0266082   .0244774    -1.09   0.277    -.0745831    .0213667
                  urban |    .108739   .0249646     4.36   0.000     .0598093    .1576687
                 income |   .0224394   .0062285     3.60   0.000     .0102319     .034647
                intnews |   .0058992   .0213919     0.28   0.783    -.0360282    .0478265
                  _cons |   .8456765    .127267     6.64   0.000     .5962377    1.095115
------------------------+----------------------------------------------------------------
                sigma_u |  .41070216
                sigma_e |  1.0578161
                    rho |  .13099532   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_judi2

. xtreg legit_judi c.lnfdi_capex##c.educlvl $C1 $C2 $C3, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,612
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0078                                         min =        184
     Between = 0.1883                                         avg =      615.1
     Overall = 0.0123                                         max =        871

                                                Wald chi2(14)     =     106.97
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0232161   .0085383     2.72   0.007     .0064812    .0399509
                educlvl |   .0502864   .0155156     3.24   0.001     .0198763    .0806965
                        |
c.lnfdi_capex#c.educlvl |  -.0071969   .0025064    -2.87   0.004    -.0121095   -.0022844
                        |
                    age |    .001934    .001108     1.75   0.081    -.0002377    .0041057
                 gender |    .047339   .0254409     1.86   0.063    -.0025243    .0972023
                  unemp |   -.051848   .0268922    -1.93   0.054    -.1045557    .0008598
                  urban |   .0857111   .0254325     3.37   0.001     .0358644    .1355578
                 income |   .0343748   .0065499     5.25   0.000     .0215372    .0472125
                intnews |   .0855098    .023982     3.57   0.000      .038506    .1325136
               religion |   .0500409   .0178982     2.80   0.005     .0149611    .0851208
                  remit |  -.0348169   .0307011    -1.13   0.257    -.0949901    .0253562
                married |  -.0402008   .0271237    -1.48   0.138    -.0933624    .0129607
               children |   .0102918   .0078337     1.31   0.189     -.005062    .0256456
              relatives |   .0365527   .0260722     1.40   0.161    -.0145478    .0876533
                  _cons |   .7782806   .0759103    10.25   0.000     .6294992     .927062
------------------------+----------------------------------------------------------------
                sigma_u |          0
                sigma_e |  1.0486414
                    rho |          0   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_judi3

. xtoprobit legit_judi c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -13941.171  
Iteration 1:   log likelihood = -13884.684  
Iteration 2:   log likelihood = -13884.683  

Refining starting values:

Grid node 0:   log likelihood = -13579.247

Fitting full model:

Iteration 0:   log likelihood = -13579.247  (not concave)
Iteration 1:   log likelihood = -13573.662  (not concave)
Iteration 2:   log likelihood = -13571.294  (not concave)
Iteration 3:   log likelihood = -13569.258  (not concave)
Iteration 4:   log likelihood = -13567.897  
Iteration 5:   log likelihood = -13554.129  
Iteration 6:   log likelihood = -13554.124  
Iteration 7:   log likelihood = -13554.124  

Random-effects ordered probit regression             Number of obs    = 10,407
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    415
                                                                  avg =  743.4
                                                                  max =    973

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     =  80.42
Log likelihood = -13554.124                          Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
             legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0165775   .0080625     2.06   0.040     .0007753    .0323798
                educlvl |   .0727078   .0149882     4.85   0.000     .0433315    .1020842
                        |
c.lnfdi_capex#c.educlvl |  -.0087733   .0023216    -3.78   0.000    -.0133235   -.0042231
                        |
                    age |   .0027179   .0009429     2.88   0.004     .0008698    .0045661
                 gender |   .0107584   .0238387     0.45   0.652    -.0359646    .0574815
                  unemp |   -.021465   .0254013    -0.85   0.398    -.0712506    .0283207
                  urban |    .106173   .0259294     4.09   0.000     .0553523    .1569938
                 income |   .0230773   .0064184     3.60   0.000     .0104974    .0356572
                intnews |   .0010368   .0222483     0.05   0.963     -.042569    .0446427
------------------------+----------------------------------------------------------------
                  /cut1 |   .0136763   .1083354                     -.1986572    .2260097
                  /cut2 |   .6589319   .1084771                      .4463208     .871543
                  /cut3 |   1.391185   .1088669                       1.17781     1.60456
------------------------+----------------------------------------------------------------
              /sigma2_u |   .1005786   .0388425                      .0471823    .2144035
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 661.12       Prob >= chibar2 = 0.0000

. est store legit_judi4

. xtreg legit_judi c.lnfdi_capex##c.educdev $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,681
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0074                                         min =        411
     Between = 0.0961                                         avg =      620.1
     Overall = 0.0097                                         max =        848

                                                Wald chi2(9)      =      66.11
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |  -.0075678   .0046583    -1.62   0.104    -.0166978    .0015623
                educdev |   .0178684   .0052007     3.44   0.001     .0076751    .0280616
                        |
c.lnfdi_capex#c.educdev |  -.0031017   .0008822    -3.52   0.000    -.0048309   -.0013726
                        |
                    age |   .0028798   .0010109     2.85   0.004     .0008984    .0048612
                 gender |   .0054202   .0245878     0.22   0.826    -.0427711    .0536115
                  unemp |  -.0122004   .0263205    -0.46   0.643    -.0637877     .039387
                  urban |   .0800363   .0275477     2.91   0.004     .0260438    .1340289
                 income |   .0328083   .0066145     4.96   0.000     .0198441    .0457724
                intnews |   .0003822   .0231746     0.02   0.987    -.0450391    .0458035
                  _cons |   1.032215   .0760055    13.58   0.000     .8832473    1.181184
------------------------+----------------------------------------------------------------
                sigma_u |  .21378726
                sigma_e |  1.0482066
                    rho |  .03993649   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_judi5

. xtreg legit_judi c.lnfdi_jobs##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,407
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0075                                         min =        415
     Between = 0.0505                                         avg =      743.4
     Overall = 0.0091                                         max =        973

                                                Wald chi2(9)      =      78.80
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
            legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0084573   .0066294     1.28   0.202    -.0045362    .0214508
               educlvl |   .0621392   .0148939     4.17   0.000     .0329476    .0913307
                       |
c.lnfdi_jobs#c.educlvl |  -.0059301   .0019566    -3.03   0.002    -.0097651   -.0020952
                       |
                   age |   .0025511   .0009065     2.81   0.005     .0007744    .0043279
                gender |    .013542   .0229754     0.59   0.556    -.0314889    .0585729
                 unemp |  -.0253894   .0244808    -1.04   0.300    -.0733708     .022592
                 urban |   .1078664    .024925     4.33   0.000     .0590144    .1567184
                income |   .0225575   .0062334     3.62   0.000     .0103402    .0347748
               intnews |   .0065126   .0213891     0.30   0.761    -.0354093    .0484346
                 _cons |   .8638084   .1274205     6.78   0.000     .6140688    1.113548
-----------------------+----------------------------------------------------------------
               sigma_u |   .4086327
               sigma_e |  1.0580615
                   rho |  .12979708   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store legit_judi6

. * Legitimacy Belief: Military Control
. xtreg legit_mili c.lnfdi_capex##c.educlvl $C1, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,749
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0094                                         min =        326
     Between = 0.1464                                         avg =      767.8
     Overall = 0.0201                                         max =      1,218

                                                Wald chi2(7)      =     103.08
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_mili | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0412797   .0064637     6.39   0.000      .028611    .0539484
                educlvl |   .0290948   .0118793     2.45   0.014     .0058117    .0523778
                        |
c.lnfdi_capex#c.educlvl |  -.0045898   .0018667    -2.46   0.014    -.0082485   -.0009312
                        |
                    age |  -.0021096   .0007529    -2.80   0.005    -.0035853   -.0006338
                 gender |   .0014797   .0191355     0.08   0.938    -.0360251    .0389846
                  unemp |    .002019   .0202563     0.10   0.921    -.0376826    .0417206
                  urban |   .0700312   .0203233     3.45   0.001     .0301982    .1098641
                  _cons |   .4702913   .1072531     4.38   0.000     .2600791    .6805034
------------------------+----------------------------------------------------------------
                sigma_u |  .34901612
                sigma_e |  .89817109
                    rho |  .13118919   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_mili1

. xtreg legit_mili c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      9,869
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0106                                         min =        307
     Between = 0.1600                                         avg =      704.9
     Overall = 0.0224                                         max =        909

                                                Wald chi2(9)      =     106.77
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_mili | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0408253   .0067398     6.06   0.000     .0276155     .054035
                educlvl |   .0310286   .0125275     2.48   0.013     .0064752    .0555819
                        |
c.lnfdi_capex#c.educlvl |  -.0050513   .0019441    -2.60   0.009    -.0088616    -.001241
                        |
                    age |  -.0024315   .0007952    -3.06   0.002    -.0039899    -.000873
                 gender |  -.0064395    .020066    -0.32   0.748    -.0457681    .0328892
                  unemp |  -.0007747   .0213437    -0.04   0.971    -.0426076    .0410583
                  urban |    .072845   .0217047     3.36   0.001     .0303046    .1153855
                 income |   .0067972   .0054249     1.25   0.210    -.0038355    .0174299
                intnews |  -.0481037   .0186852    -2.57   0.010     -.084726   -.0114815
                  _cons |   .5193063   .1237964     4.19   0.000     .2766698    .7619427
------------------------+----------------------------------------------------------------
                sigma_u |  .41236528
                sigma_e |  .89937452
                    rho |  .17370693   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_mili2

. xtreg legit_mili c.lnfdi_capex##c.educlvl $C1 $C2 $C3, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,203
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0057                                         min =        175
     Between = 0.2845                                         avg =      585.9
     Overall = 0.0278                                         max =        865

                                                Wald chi2(14)     =     234.28
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_mili | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0355129   .0075152     4.73   0.000     .0207834    .0502424
                educlvl |   .0307677   .0137168     2.24   0.025     .0038833    .0576521
                        |
c.lnfdi_capex#c.educlvl |  -.0039926   .0022105    -1.81   0.071     -.008325    .0003398
                        |
                    age |  -.0022216   .0009836    -2.26   0.024    -.0041495   -.0002936
                 gender |   .0587798   .0225477     2.61   0.009     .0145871    .1029724
                  unemp |    -.06985   .0238127    -2.93   0.003     -.116522   -.0231779
                  urban |   .0826385   .0225567     3.66   0.000     .0384283    .1268488
                 income |  -.0059527   .0058016    -1.03   0.305    -.0173236    .0054182
                intnews |  -.0073701   .0212596    -0.35   0.729    -.0490381    .0342979
               religion |   .0775795   .0158144     4.91   0.000     .0465838    .1085753
                  remit |   .0619744   .0269587     2.30   0.022     .0091364    .1148124
                married |  -.0557738   .0241122    -2.31   0.021    -.1030328   -.0085147
               children |  -.0114043   .0069295    -1.65   0.100    -.0249859    .0021773
              relatives |   .0730296   .0230861     3.16   0.002     .0277816    .1182775
                  _cons |   .4372809   .0671836     6.51   0.000     .3056034    .5689583
------------------------+----------------------------------------------------------------
                sigma_u |          0
                sigma_e |  .89156906
                    rho |          0   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_mili3

. xtoprobit legit_mili c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood =  -10999.83  
Iteration 1:   log likelihood = -10862.409  
Iteration 2:   log likelihood = -10862.227  
Iteration 3:   log likelihood = -10862.227  

Refining starting values:

Grid node 0:   log likelihood = -10446.699

Fitting full model:

Iteration 0:   log likelihood = -10446.699  (not concave)
Iteration 1:   log likelihood = -10442.918  (not concave)
Iteration 2:   log likelihood =  -10439.25  
Iteration 3:   log likelihood = -10437.327  (backed up)
Iteration 4:   log likelihood =  -10414.07  
Iteration 5:   log likelihood =  -10414.04  
Iteration 6:   log likelihood =  -10414.04  

Random-effects ordered probit regression             Number of obs    =  9,869
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    307
                                                                  avg =  704.9
                                                                  max =    909

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     = 122.35
Log likelihood = -10414.04                           Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
             legit_mili | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0652478   .0093858     6.95   0.000     .0468519    .0836436
                educlvl |   .0646933   .0174157     3.71   0.000     .0305592    .0988275
                        |
c.lnfdi_capex#c.educlvl |  -.0099267   .0026571    -3.74   0.000    -.0151346   -.0047189
                        |
                    age |  -.0031564    .001053    -3.00   0.003    -.0052202   -.0010925
                 gender |  -.0055448   .0261464    -0.21   0.832    -.0567908    .0457011
                  unemp |   .0030803   .0279602     0.11   0.912    -.0517206    .0578812
                  urban |   .1089327   .0285347     3.82   0.000     .0530057    .1648597
                 income |   .0076651   .0070853     1.08   0.279    -.0062219    .0215521
                intnews |  -.0676308   .0244826    -2.76   0.006    -.1156159   -.0196457
------------------------+----------------------------------------------------------------
                  /cut1 |   .5284298   .1326227                      .2684941    .7883654
                  /cut2 |   1.270035   .1329637                      1.009431     1.53064
                  /cut3 |   1.888316   .1335663                      1.626531    2.150101
------------------------+----------------------------------------------------------------
              /sigma2_u |   .1622754   .0625476                      .0762368    .3454146
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 896.37       Prob >= chibar2 = 0.0000

. est store legit_mili4

. xtreg legit_mili c.lnfdi_capex##c.educdev $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      8,249
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0064                                         min =        303
     Between = 0.1992                                         avg =      589.2
     Overall = 0.0182                                         max =        817

                                                Wald chi2(9)      =      54.09
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_mili | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |     .01856   .0041517     4.47   0.000     .0104228    .0266973
                educdev |   .0096689    .004585     2.11   0.035     .0006824    .0186555
                        |
c.lnfdi_capex#c.educdev |  -.0018619   .0007782    -2.39   0.017    -.0033871   -.0003367
                        |
                    age |  -.0026194   .0008962    -2.92   0.003     -.004376   -.0008629
                 gender |  -.0074394   .0216903    -0.34   0.732    -.0499517    .0350728
                  unemp |  -.0208788    .023185    -0.90   0.368    -.0663206     .024563
                  urban |   .0614925   .0242494     2.54   0.011     .0139646    .1090204
                 income |   .0087233   .0058285     1.50   0.134    -.0027004     .020147
                intnews |  -.0208369   .0204237    -1.02   0.308    -.0608665    .0191928
                  _cons |   .6622746   .1057239     6.26   0.000     .4550595    .8694896
------------------------+----------------------------------------------------------------
                sigma_u |   .3589891
                sigma_e |  .90125329
                    rho |  .13693441   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. est store legit_mili5

. xtreg legit_mili c.lnfdi_jobs##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      9,869
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0103                                         min =        307
     Between = 0.1456                                         avg =      704.9
     Overall = 0.0211                                         max =        909

                                                Wald chi2(9)      =     103.63
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

----------------------------------------------------------------------------------------
            legit_mili | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
            lnfdi_jobs |   .0315355   .0057848     5.45   0.000     .0201975    .0428734
               educlvl |   .0252832   .0130163     1.94   0.052    -.0002282    .0507946
                       |
c.lnfdi_jobs#c.educlvl |  -.0031469   .0017081    -1.84   0.065    -.0064947    .0002009
                       |
                   age |   -.002369   .0007952    -2.98   0.003    -.0039277   -.0008104
                gender |   -.006649   .0200679    -0.33   0.740    -.0459815    .0326834
                 unemp |    -.00104   .0213459    -0.05   0.961    -.0428773    .0407973
                 urban |   .0749091   .0216766     3.46   0.001     .0324237    .1173944
                income |   .0065168   .0054287     1.20   0.230    -.0041232    .0171569
               intnews |  -.0488242   .0186798    -2.61   0.009    -.0854359   -.0122126
                 _cons |   .5287582   .1223789     4.32   0.000        .2889    .7686164
-----------------------+----------------------------------------------------------------
               sigma_u |  .40422174
               sigma_e |  .89951439
                   rho |  .16801196   (fraction of variance due to u_i)
----------------------------------------------------------------------------------------

. est store legit_mili6

. 
. *** TABLE A-16: Perceived Regime Legitimacy - Main Models
. esttab legit_free2 legit_free3 legit_free4 legit_judi2 legit_judi3 legit_judi4 legit_mili2 legit_mili3 legit_mili4
> , ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl lnfdi_capex c.lnfdi_capex#c.educlvl age gender unemp urban income intnews ///
>                 religion remit married children relatives) ///
>         drop(_cons) eqlabels("" "") nogaps

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
               legit_free      legit_free      legit_free      legit_judi      legit_judi      legit_judi      legit
> _mili      legit_mili      legit_mili   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                                                                                                                    
>                                         
educlvl             0.059***        0.028**         0.077***        0.068***        0.050***        0.073***        
> 0.031**         0.031**         0.065***
                   (0.01)          (0.01)          (0.02)          (0.01)          (0.02)          (0.01)          (
> 0.01)          (0.01)          (0.02)   
lnfdi_capex         0.040***        0.062***        0.051***        0.014*          0.023***        0.017**         
> 0.041***        0.036***        0.065***
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
c.lnfdi_ca~l       -0.011***       -0.013***       -0.014***       -0.008***       -0.007***       -0.009***       -
> 0.005***       -0.004*         -0.010***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
age                -0.001          -0.004***       -0.001           0.003***        0.002*          0.003***       -
> 0.002***       -0.002**        -0.003***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
gender              0.004           0.012           0.006           0.014           0.047*          0.011          -
> 0.006           0.059***       -0.006   
                   (0.02)          (0.02)          (0.02)          (0.02)          (0.03)          (0.02)          (
> 0.02)          (0.02)          (0.03)   
unemp               0.020           0.037           0.020          -0.027          -0.052*         -0.021          -
> 0.001          -0.070***        0.003   
                   (0.02)          (0.02)          (0.03)          (0.02)          (0.03)          (0.03)          (
> 0.02)          (0.02)          (0.03)   
urban               0.030           0.000           0.038           0.109***        0.086***        0.106***        
> 0.073***        0.083***        0.109***
                   (0.02)          (0.02)          (0.03)          (0.02)          (0.03)          (0.03)          (
> 0.02)          (0.02)          (0.03)   
income              0.004           0.024***        0.006           0.022***        0.034***        0.023***        
> 0.007          -0.006           0.008   
                   (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)          (
> 0.01)          (0.01)          (0.01)   
intnews            -0.001           0.071***        0.007           0.006           0.086***        0.001          -
> 0.048**        -0.007          -0.068***
                   (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
religion                           -0.039**                                         0.050***                        
>                 0.078***                
                                   (0.02)                                          (0.02)                           
>                (0.02)                   
remit                               0.041                                          -0.035                           
>                 0.062**                 
                                   (0.03)                                          (0.03)                           
>                (0.03)                   
married                            -0.009                                          -0.040                           
>                -0.056**                 
                                   (0.02)                                          (0.03)                           
>                (0.02)                   
children                            0.016**                                         0.010                           
>                -0.011*                  
                                   (0.01)                                          (0.01)                           
>                (0.01)                   
relatives                           0.071***                                        0.037                           
>                 0.073***                
                                   (0.02)                                          (0.03)                           
>                (0.02)                   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        10480            8686           10480           10407            8612           10407           
>  9869            8203            9869   
\# of coun~s           14              14              14              14              14              14           
>    14              14              14   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

.         
. *** TABLE A-17: Perceived Regime Legitimacy - Robustness Tests
. esttab legit_free1 legit_free5 legit_free6 legit_judi1 legit_judi5 legit_judi6 legit_mili1 legit_mili5 legit_mili6
> , ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl educdev lnfdi_capex c.lnfdi_capex#c.educlvl lnfdi_jobs c.lnfdi_jobs#c.educlvl c.lnfdi_capex#
> c.educdev ///
>                 age gender unemp urban income intnews) ///
>         drop(_cons) eqlabels("" "") nogaps

--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)           
>   (7)             (8)             (9)   
               legit_free      legit_free      legit_free      legit_judi      legit_judi      legit_judi      legit
> _mili      legit_mili      legit_mili   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
educlvl             0.052***                        0.058***        0.068***                        0.062***        
> 0.029**                         0.025*  
                   (0.01)                          (0.01)          (0.01)                          (0.01)          (
> 0.01)                          (0.01)   
educdev                             0.015***                                        0.018***                        
>                 0.010**                 
                                   (0.00)                                          (0.01)                           
>                (0.00)                   
lnfdi_capex         0.038***        0.003                           0.017**        -0.008                           
> 0.041***        0.019***                
                   (0.01)          (0.00)                          (0.01)          (0.00)                          (
> 0.01)          (0.00)                   
c.lnfdi_ca~l       -0.010***                                       -0.007***                                       -
> 0.005**                                 
                   (0.00)                                          (0.00)                                          (
> 0.00)                                   
lnfdi_jobs                                          0.033***                                        0.008           
>                                 0.032***
                                                   (0.01)                                          (0.01)           
>                                (0.01)   
c.lnfdi_jo~l                                       -0.009***                                       -0.006***        
>                                -0.003*  
                                                   (0.00)                                          (0.00)           
>                                (0.00)   
c.lnfdi_ca~v                       -0.003***                                       -0.003***                        
>                -0.002**                 
                                   (0.00)                                          (0.00)                           
>                (0.00)                   
age                -0.001          -0.001*         -0.001           0.002***        0.003***        0.003***       -
> 0.002***       -0.003***       -0.002***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (
> 0.00)          (0.00)          (0.00)   
gender              0.002           0.026           0.004           0.014           0.005           0.014           
> 0.001          -0.007          -0.007   
                   (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
unemp               0.024           0.008           0.021          -0.031          -0.012          -0.025           
> 0.002          -0.021          -0.001   
                   (0.02)          (0.02)          (0.02)          (0.02)          (0.03)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
urban               0.030           0.045*          0.031           0.119***        0.080***        0.108***        
> 0.070***        0.061**         0.075***
                   (0.02)          (0.02)          (0.02)          (0.02)          (0.03)          (0.02)          (
> 0.02)          (0.02)          (0.02)   
income                              0.011*          0.004                           0.033***        0.023***        
>                 0.009           0.007   
                                   (0.01)          (0.01)                          (0.01)          (0.01)           
>                (0.01)          (0.01)   
intnews                             0.016          -0.001                           0.000           0.007           
>                -0.021          -0.049***
                                   (0.02)          (0.02)                          (0.02)          (0.02)           
>                (0.02)          (0.02)   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
\# respond~s        11442            8752           10480           11365            8681           10407           
> 10749            8249            9869   
\# of coun~s           14              14              14              14              14              14           
>    14              14              14   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000           
> 0.000           0.000           0.000   
--------------------------------------------------------------------------------------------------------------------
> ----------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** TABLE 3: FDI and Political Preferences, Pew Global Attitudes
. esttab problem2 worse2 leadsat2 legit_free2 legit_judi2 legit_mili2, ///
>         b(3) se(2) starlevels(* 0.10 ** 0.05 *** 0.01) ///
>         stats(N N_g p, fmt(0 0 3) labels("\# respondents" "\# of countries" "Prob > Chi2")) ///
>         order(educlvl lnfdi_capex c.lnfdi_capex#c.educlvl age gender unemp urban income intnews) ///
>         drop(_cons) eqlabels("" "") nogaps

------------------------------------------------------------------------------------------------------------
                      (1)             (2)             (3)             (4)             (5)             (6)   
                  problem           worse         leadsat      legit_free      legit_judi      legit_mili   
------------------------------------------------------------------------------------------------------------
educlvl             0.010           0.057**         0.030**         0.059***        0.068***        0.031** 
                   (0.01)          (0.03)          (0.01)          (0.01)          (0.01)          (0.01)   
lnfdi_capex         0.006*          0.013          -0.009           0.040***        0.014*          0.041***
                   (0.00)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)   
c.lnfdi_ca~l       -0.003***       -0.008*         -0.004**        -0.011***       -0.008***       -0.005***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)   
age                -0.002***        0.011***       -0.001          -0.001           0.003***       -0.002***
                   (0.00)          (0.00)          (0.00)          (0.00)          (0.00)          (0.00)   
gender             -0.014           0.129***        0.012           0.004           0.014          -0.006   
                   (0.01)          (0.04)          (0.02)          (0.02)          (0.02)          (0.02)   
unemp               0.025**        -0.047           0.004           0.020          -0.027          -0.001   
                   (0.01)          (0.05)          (0.02)          (0.02)          (0.02)          (0.02)   
urban               0.051***       -0.052           0.089***        0.030           0.109***        0.073***
                   (0.01)          (0.05)          (0.02)          (0.02)          (0.02)          (0.02)   
income             -0.025***        0.011          -0.012**         0.004           0.022***        0.007   
                   (0.00)          (0.01)          (0.01)          (0.01)          (0.01)          (0.01)   
intnews             0.025***       -0.096**         0.029          -0.001           0.006          -0.048** 
                   (0.01)          (0.04)          (0.02)          (0.02)          (0.02)          (0.02)   
------------------------------------------------------------------------------------------------------------
\# respond~s        10646            9661           10058           10480           10407            9869   
\# of coun~s           14              14              13              14              14              14   
Prob > Chi2         0.000           0.000           0.000           0.000           0.000           0.000   
------------------------------------------------------------------------------------------------------------
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01

. 
. *** Self-Reported Economic Problems
. xtreg problem c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,646
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0134                                         min =        446
     Between = 0.0692                                         avg =      760.4
     Overall = 0.0145                                         max =      1,029

                                                Wald chi2(9)      =     145.18
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0059443   .0032836     1.81   0.070    -.0004915      .01238
                educlvl |   .0099307   .0060936     1.63   0.103    -.0020126     .021874
                        |
c.lnfdi_capex#c.educlvl |  -.0025876   .0009506    -2.72   0.006    -.0044507   -.0007244
                        |
                    age |  -.0015805     .00039    -4.05   0.000    -.0023449   -.0008161
                 gender |  -.0139884    .009899    -1.41   0.158      -.03339    .0054132
                  unemp |   .0252284   .0105306     2.40   0.017     .0045888     .045868
                  urban |   .0511622    .010663     4.80   0.000     .0302631    .0720613
                 income |  -.0246401   .0026849    -9.18   0.000    -.0299025   -.0193777
                intnews |    .024564    .009193     2.67   0.008      .006546    .0425819
                  _cons |   .3219699   .0345449     9.32   0.000     .2542631    .3896767
------------------------+----------------------------------------------------------------
                sigma_u |  .07879719
                sigma_e |   .4605604
                    rho |  .02843925   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_capex) at(educlvl=(0(1)5)) vsquish

Average marginal effects                                Number of obs = 10,646
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_capex
1._at: educlvl = 0
2._at: educlvl = 1
3._at: educlvl = 2
4._at: educlvl = 3
5._at: educlvl = 4
6._at: educlvl = 5

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_capex  |
         _at |
          1  |   .0059443   .0032836     1.81   0.070    -.0004915      .01238
          2  |   .0033567   .0025439     1.32   0.187    -.0016292    .0083426
          3  |   .0007691   .0019919     0.39   0.699     -.003135    .0046732
          4  |  -.0018185   .0018088    -1.01   0.315    -.0053636    .0017266
          5  |  -.0044061   .0020935    -2.10   0.035    -.0085092   -.0003029
          6  |  -.0069936   .0027021    -2.59   0.010    -.0122896   -.0016977
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Self-Reported Economic Problems", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(-0.5 5.5)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(pew_problem, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(, labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. 
. *** Legitimacy Belief: Freedom of Speech
. xtreg legit_free c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,480
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0044                                         min =        416
     Between = 0.0513                                         avg =      748.6
     Overall = 0.0072                                         max =        946

                                                Wald chi2(9)      =      46.35
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0403298   .0067375     5.99   0.000     .0271246     .053535
                educlvl |   .0589642   .0124389     4.74   0.000     .0345843     .083344
                        |
c.lnfdi_capex#c.educlvl |  -.0108482   .0019428    -5.58   0.000    -.0146561   -.0070404
                        |
                    age |  -.0012338   .0007944    -1.55   0.120    -.0027909    .0003232
                 gender |   .0041273    .020122     0.21   0.837    -.0353111    .0435658
                  unemp |   .0204996   .0214193     0.96   0.339    -.0214815    .0624806
                  urban |   .0297488   .0218841     1.36   0.174    -.0131432    .0726409
                 income |   .0035284   .0054681     0.65   0.519    -.0071888    .0142457
                intnews |  -.0012428   .0187543    -0.07   0.947    -.0380005    .0355149
                  _cons |    .635365   .1043603     6.09   0.000     .4308225    .8399075
------------------------+----------------------------------------------------------------
                sigma_u |   .3293814
                sigma_e |  .93057463
                    rho |  .11133547   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_capex) at(educlvl=(0(1)5)) vsquish

Average marginal effects                                Number of obs = 10,480
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_capex
1._at: educlvl = 0
2._at: educlvl = 1
3._at: educlvl = 2
4._at: educlvl = 3
5._at: educlvl = 4
6._at: educlvl = 5

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_capex  |
         _at |
          1  |   .0403298   .0067375     5.99   0.000     .0271246     .053535
          2  |   .0294816   .0052277     5.64   0.000     .0192355    .0397277
          3  |   .0186334   .0041004     4.54   0.000     .0105967      .02667
          4  |   .0077852   .0037211     2.09   0.036     .0004919    .0150784
          5  |   -.003063   .0042929    -0.71   0.476     -.011477    .0053509
          6  |  -.0139113   .0055282    -2.52   0.012    -.0247463   -.0030763
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Legitimacy Belief: Freedom of Speech", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(-0.5 5.5)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(pew_legit_free, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(, labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. 
. *** Legitimacy Belief: Judiciary
. xtreg legit_judi c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =     10,407
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0080                                         min =        415
     Between = 0.0274                                         avg =      743.4
     Overall = 0.0089                                         max =        973

                                                Wald chi2(9)      =      83.57
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
             legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0137792   .0077234     1.78   0.074    -.0013584    .0289168
                educlvl |   .0681143   .0143506     4.75   0.000     .0399877    .0962409
                        |
c.lnfdi_capex#c.educlvl |  -.0081975   .0022278    -3.68   0.000    -.0125639   -.0038311
                        |
                    age |   .0025428   .0009064     2.81   0.005     .0007663    .0043193
                 gender |   .0143746   .0229721     0.63   0.531    -.0306498     .059399
                  unemp |  -.0266082   .0244774    -1.09   0.277    -.0745831    .0213667
                  urban |    .108739   .0249646     4.36   0.000     .0598093    .1576687
                 income |   .0224394   .0062285     3.60   0.000     .0102319     .034647
                intnews |   .0058992   .0213919     0.28   0.783    -.0360282    .0478265
                  _cons |   .8456765    .127267     6.64   0.000     .5962377    1.095115
------------------------+----------------------------------------------------------------
                sigma_u |  .41070216
                sigma_e |  1.0578161
                    rho |  .13099532   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. margins, dydx(lnfdi_capex) at(educlvl=(0(1)5)) vsquish

Average marginal effects                                Number of obs = 10,407
Model VCE: Conventional

Expression: Linear prediction, predict()
dy/dx wrt:  lnfdi_capex
1._at: educlvl = 0
2._at: educlvl = 1
3._at: educlvl = 2
4._at: educlvl = 3
5._at: educlvl = 4
6._at: educlvl = 5

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
lnfdi_capex  |
         _at |
          1  |   .0137792   .0077234     1.78   0.074    -.0013584    .0289168
          2  |   .0055817   .0059915     0.93   0.352    -.0061615    .0173248
          3  |  -.0026159   .0046981    -0.56   0.578     -.011824    .0065923
          4  |  -.0108134   .0042629    -2.54   0.011    -.0191685   -.0024583
          5  |  -.0190109   .0049192    -3.86   0.000    -.0286524   -.0093695
          6  |  -.0272085   .0063364    -4.29   0.000    -.0396277   -.0147893
------------------------------------------------------------------------------

. marginsplot, level(95) ///
>         plotopts(msymbol(i) lcolor(black)) ///
>         ciopts(recast(rline) lpattern(dash) lcolor(black)) ///
>         title("Legitimacy Belief: Judiciary", size(medium) margin(medsmall)) ///
>         xtitle("Individual education level", size(medsmall) margin(medsmall)) ///
>         ytitle("Marginal effect of FDI exposure", size(medsmall) margin(off)) ///
>         ylabel(, nogrid)  ///
>         xscale(range(-0.5 5.5)) ///
>         xlabel(, labsize(small) nogrid) ///
>         yline(0, lcolor(black) lwidth(.2)) ///
>         graphregion(fcolor(white) ilcolor(white) lcolor(white)) ///
>         name(pew_legit_judi, replace) xsize(2) ysize(2) ///
>         addplot(hist educlvl, disc yaxis(2) legend(off) ///
>                 ytitle("", axis(2)) ylabel(none, axis(2)) ///
>                 xlabel(, labsize(small) nogrid) ///
>                 fcolor(none) lcolor(gs12))

Variables that uniquely identify margins: educlvl
(note:  named style off not found in class margin, default attributes used)

. 
. *** FIGURE A-5: Combined marginal effects plots
. graph combine pew_problem pew_legit_free pew_legit_judi, col(1) iscale(.8) ///
>         xsize(2) ysize(5) name(pew_alleffects, replace)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)
(note:  named style off not found in class margin, default attributes used)

. 
. *** TABLE A-18: Substantial Effect Sizes - All Models
. * Self-Reported Economic Problems
. xtprobit problem c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -6781.4016  
Iteration 1:   log likelihood = -6676.1116  
Iteration 2:   log likelihood = -6676.0172  
Iteration 3:   log likelihood = -6676.0172  

Fitting full model:

rho =  0.0     log likelihood = -6676.0172
rho =  0.1     log likelihood = -6576.8494
rho =  0.2     log likelihood = -6585.4848

Iteration 0:   log likelihood = -6567.8651  
Iteration 1:   log likelihood = -6545.5394  
Iteration 2:   log likelihood = -6544.4658  
Iteration 3:   log likelihood = -6544.4549  
Iteration 4:   log likelihood = -6544.4549  

Random-effects probit regression                     Number of obs    = 10,646
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    446
                                                                  avg =  760.4
                                                                  max =  1,029

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     = 140.74
Log likelihood = -6544.4549                          Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
                problem | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0163979   .0092753     1.77   0.077    -.0017814    .0345771
                educlvl |    .031102   .0173633     1.79   0.073    -.0029295    .0651335
                        |
c.lnfdi_capex#c.educlvl |  -.0076077   .0027026    -2.81   0.005    -.0129047   -.0023108
                        |
                    age |   -.004669   .0011241    -4.15   0.000    -.0068722   -.0024658
                 gender |  -.0393389   .0284812    -1.38   0.167     -.095161    .0164832
                  unemp |   .0698041   .0301386     2.32   0.021     .0107335    .1288746
                  urban |   .1408756   .0302584     4.66   0.000     .0815702     .200181
                 income |  -.0715627   .0079155    -9.04   0.000    -.0870768   -.0560486
                intnews |   .0697817   .0263099     2.65   0.008     .0182152    .1213481
                  _cons |  -.4771795   .1045628    -4.56   0.000    -.6821187   -.2722403
------------------------+----------------------------------------------------------------
               /lnsig2u |  -2.697409   .3983578                     -3.478176   -1.916642
------------------------+----------------------------------------------------------------
                sigma_u |   .2595763   .0517021                      .1756805    .3835362
                    rho |   .0631264   .0235595                      .0299396    .1282364
-----------------------------------------------------------------------------------------
LR test of rho=0: chibar2(01) = 263.12                 Prob >= chibar2 = 0.000

. margins, post ///
>         at(educlvl=1.40 lnfdi_capex=2.08) ///
>         at(educlvl=1.40 lnfdi_capex=8.42) ///
>         at(educlvl=4.50 lnfdi_capex=2.08) ///
>         at(educlvl=4.50 lnfdi_capex=8.42) atmeans

Adjusted predictions                                    Number of obs = 10,646
Model VCE: OIM

Expression: Pr(problem=1), predict(pr)
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .3210304   .0260621    12.32   0.000     .2699496    .3721112
          2  |   .3337608   .0266808    12.51   0.000     .2814674    .3860542
          3  |   .3376164   .0268617    12.57   0.000     .2849685    .3902643
          4  |   .2985984   .0252769    11.81   0.000     .2490567    .3481402
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 10,646
Model VCE: OIM

Expression: Pr(problem=1), predict(pr)
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.23502 (mean)
       gender      = .4828104 (mean)
       unemp       = .4065377 (mean)
       urban       = .5750517 (mean)
       income      = .2891227 (mean)
       intnews     = .5197257 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .3210304  _b[1bn._at]
          2  |   .3337608  _b[2._at]
          3  |   .3376164  _b[3._at]
          4  |   .2985984  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0127304   .0144295     0.88   0.378    -.0155508    .0410116
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    0.78
         Prob > chi2 =    0.3776

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   -.039018   .0150829    -2.59   0.010      -.06858    -.009456
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    6.69
         Prob > chi2 =    0.0097

. * Expected Social Decline
. xtreg worse c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Random-effects GLS regression                   Number of obs     =      9,661
Group variable: ccode                           Number of groups  =         14

R-squared:                                      Obs per group:
     Within  = 0.0064                                         min =        365
     Between = 0.1297                                         avg =      690.1
     Overall = 0.0090                                         max =        886

                                                Wald chi2(9)      =      62.86
corr(u_i, X) = 0 (assumed)                      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------
                  worse | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0132103   .0146175     0.90   0.366    -.0154394      .04186
                educlvl |    .056551   .0271405     2.08   0.037     .0033566    .1097454
                        |
c.lnfdi_capex#c.educlvl |  -.0082276   .0042072    -1.96   0.051    -.0164736    .0000183
                        |
                    age |   .0110923    .001734     6.40   0.000     .0076937    .0144908
                 gender |   .1288678   .0437296     2.95   0.003     .0431593    .2145763
                  unemp |  -.0470409   .0464929    -1.01   0.312    -.1381653    .0440835
                  urban |  -.0517529   .0475957    -1.09   0.277    -.1450387    .0415329
                 income |   .0112401   .0117153     0.96   0.337    -.0117216    .0342017
                intnews |  -.0963971   .0406422    -2.37   0.018    -.1760543     -.01674
                  _cons |  -1.833556   .1750198   -10.48   0.000    -2.176589   -1.490524
------------------------+----------------------------------------------------------------
                sigma_u |  .46837985
                sigma_e |   1.937405
                    rho |  .05521878   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. margins, post ///
>         at(educlvl=1.40 lnfdi_capex=2.08) ///
>         at(educlvl=1.40 lnfdi_capex=8.42) ///
>         at(educlvl=4.50 lnfdi_capex=2.08) ///
>         at(educlvl=4.50 lnfdi_capex=8.42) atmeans

Adjusted predictions                                     Number of obs = 9,661
Model VCE: Conventional

Expression: Linear prediction, predict()
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |  -1.396683   .1330255   -10.50   0.000    -1.657408   -1.135958
          2  |  -1.385958   .1341187   -10.33   0.000    -1.648826    -1.12309
          3  |  -1.274427   .1336949    -9.53   0.000    -1.536464    -1.01239
          4  |  -1.425408   .1326611   -10.74   0.000    -1.685419   -1.165397
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                     Number of obs = 9,661
Model VCE: Conventional

Expression: Linear prediction, predict()
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.01522 (mean)
       gender      = .4812131 (mean)
       unemp       =  .409999 (mean)
       urban       = .5905186 (mean)
       income      = .3291585 (mean)
       intnews     = .5195114 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |  -1.396683  _b[1bn._at]
          2  |  -1.385958  _b[2._at]
          3  |  -1.274427  _b[3._at]
          4  |  -1.425408  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0107248   .0648823     0.17   0.869    -.1164422    .1378918
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    0.03
         Prob > chi2 =    0.8687

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.1509812   .0666243    -2.27   0.023    -.2815624   -.0204001
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    5.14
         Prob > chi2 =    0.0234

. * Leader Satisfaction
. xtoprobit leadsat c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -12681.668  
Iteration 1:   log likelihood = -12611.512  
Iteration 2:   log likelihood = -12611.495  
Iteration 3:   log likelihood = -12611.495  

Refining starting values:

Grid node 0:   log likelihood = -11718.156

Fitting full model:

Iteration 0:   log likelihood = -11718.156  (not concave)
Iteration 1:   log likelihood = -11714.383  (not concave)
Iteration 2:   log likelihood = -11710.623  (not concave)
Iteration 3:   log likelihood =  -11709.88  
Iteration 4:   log likelihood = -11614.413  
Iteration 5:   log likelihood =  -11578.68  
Iteration 6:   log likelihood = -11578.387  
Iteration 7:   log likelihood = -11578.376  
Iteration 8:   log likelihood = -11578.376  

Random-effects ordered probit regression             Number of obs    = 10,058
Group variable: ccode                                Number of groups =     13

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    568
                                                                  avg =  773.7
                                                                  max =    982

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     =  53.21
Log likelihood = -11578.376                          Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
                leadsat | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |  -.0063723   .0083766    -0.76   0.447    -.0227901    .0100456
                educlvl |     .04037   .0156299     2.58   0.010     .0097359     .071004
                        |
c.lnfdi_capex#c.educlvl |  -.0053062   .0024232    -2.19   0.029    -.0100556   -.0005568
                        |
                    age |  -.0011824   .0009777    -1.21   0.226    -.0030986    .0007338
                 gender |   .0194437   .0247081     0.79   0.431    -.0289832    .0678706
                  unemp |   .0125304   .0262516     0.48   0.633    -.0389219    .0639826
                  urban |   .1205257   .0265122     4.55   0.000     .0685627    .1724887
                 income |  -.0138091   .0066044    -2.09   0.037    -.0267534   -.0008648
                intnews |   .0271915    .023112     1.18   0.239    -.0181072    .0724903
------------------------+----------------------------------------------------------------
                  /cut1 |  -.2291822   .1730844                     -.5684214    .1100571
                  /cut2 |   .8278398   .1732208                      .4883333    1.167346
                  /cut3 |   1.468959   .1735063                      1.128893    1.809025
------------------------+----------------------------------------------------------------
              /sigma2_u |   .3258566   .1287053                      .1502529    .7066917
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 2066.24      Prob >= chibar2 = 0.0000

. margins, post predict(outcome(4)) ///
>         at(educlvl=1.40 lnfdi_capex=2.08) ///
>         at(educlvl=1.40 lnfdi_capex=8.42) ///
>         at(educlvl=4.50 lnfdi_capex=2.08) ///
>         at(educlvl=4.50 lnfdi_capex=8.42) atmeans

Adjusted predictions                                    Number of obs = 10,058
Model VCE: OIM

Expression: Pr(4.leadsat), predict(outcome(4))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .1143739   .0293002     3.90   0.000     .0569465    .1718013
          2  |   .1003465     .02688     3.73   0.000     .0476627    .1530304
          3  |   .1303752   .0318616     4.09   0.000     .0679277    .1928228
          4  |   .0983209   .0264806     3.71   0.000     .0464199    .1502219
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 10,058
Model VCE: OIM

Expression: Pr(4.leadsat), predict(outcome(4))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.41241 (mean)
       gender      =  .483098 (mean)
       unemp       = .4101213 (mean)
       urban       = .6039968 (mean)
       income      = .3207397 (mean)
       intnews     = .5234639 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .1143739  _b[1bn._at]
          2  |   .1003465  _b[2._at]
          3  |   .1303752  _b[3._at]
          4  |   .0983209  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0140274   .0064591    -2.17   0.030     -.026687   -.0013678
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    4.72
         Prob > chi2 =    0.0299

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0320543   .0085317    -3.76   0.000    -.0487762   -.0153325
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =   14.12
         Prob > chi2 =    0.0002

. * Legitimacy Belief: Freedom of Speech
. xtoprobit legit_free c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -12752.828  
Iteration 1:   log likelihood = -12691.103  
Iteration 2:   log likelihood = -12691.096  
Iteration 3:   log likelihood = -12691.096  

Refining starting values:

Grid node 0:   log likelihood = -12216.297

Fitting full model:

Iteration 0:   log likelihood = -12216.297  (not concave)
Iteration 1:   log likelihood = -12211.824  (not concave)
Iteration 2:   log likelihood = -12209.787  (not concave)
Iteration 3:   log likelihood = -12208.109  
Iteration 4:   log likelihood = -12199.841  
Iteration 5:   log likelihood = -12175.222  
Iteration 6:   log likelihood =   -12175.2  
Iteration 7:   log likelihood =   -12175.2  

Random-effects ordered probit regression             Number of obs    = 10,480
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    416
                                                                  avg =  748.6
                                                                  max =    946

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     =  49.37
Log likelihood = -12175.2                            Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
             legit_free | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0512057   .0082324     6.22   0.000     .0350704    .0673409
                educlvl |   .0770005   .0152234     5.06   0.000     .0471632    .1068379
                        |
c.lnfdi_capex#c.educlvl |  -.0138897   .0023752    -5.85   0.000     -.018545   -.0092345
                        |
                    age |  -.0013382   .0009695    -1.38   0.168    -.0032384    .0005621
                 gender |   .0063591   .0244672     0.26   0.795    -.0415957     .054314
                  unemp |   .0198601   .0259762     0.76   0.445    -.0310524    .0707726
                  urban |   .0375057   .0264462     1.42   0.156    -.0143279    .0893393
                 income |   .0058136    .006595     0.88   0.378    -.0071123    .0187395
                intnews |   .0069949   .0228334     0.31   0.759    -.0377576    .0517475
------------------------+----------------------------------------------------------------
                  /cut1 |   .2453999   .1285916                      -.006635    .4974348
                  /cut2 |   1.020493   .1288377                      .7679755     1.27301
                  /cut3 |   1.764555   .1294513                      1.510835    2.018274
------------------------+----------------------------------------------------------------
              /sigma2_u |   .1652248   .0633916                      .0778925    .3504732
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 1031.79      Prob >= chibar2 = 0.0000

. margins, post predict(outcome(3)) ///
>         at(educlvl=1.40 lnfdi_capex=2.08) ///
>         at(educlvl=1.40 lnfdi_capex=8.42) ///
>         at(educlvl=4.50 lnfdi_capex=2.08) ///
>         at(educlvl=4.50 lnfdi_capex=8.42) atmeans

Adjusted predictions                                    Number of obs = 10,480
Model VCE: OIM

Expression: Pr(3.legit_free), predict(outcome(3))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .0691885   .0148248     4.67   0.000     .0401324    .0982445
          2  |   .0976041     .01894     5.15   0.000     .0604823    .1347258
          3  |   .0895228   .0178691     5.01   0.000        .0545    .1245456
          4  |   .0792615   .0163555     4.85   0.000     .0472053    .1113177
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 10,480
Model VCE: OIM

Expression: Pr(3.legit_free), predict(outcome(3))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.24027 (mean)
       gender      = .4792939 (mean)
       unemp       =  .403626 (mean)
       urban       = .5802481 (mean)
       income      =  .303626 (mean)
       intnews     = .5222328 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .0691885  _b[1bn._at]
          2  |   .0976041  _b[2._at]
          3  |   .0895228  _b[3._at]
          4  |   .0792615  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0284156   .0066051     4.30   0.000     .0154699    .0413613
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =   18.51
         Prob > chi2 =    0.0000

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0102613     .00556    -1.85   0.065    -.0211587    .0006361
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    3.41
         Prob > chi2 =    0.0650

. * Legitimacy Belief: Judiciary
. xtoprobit legit_judi c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood = -13941.171  
Iteration 1:   log likelihood = -13884.684  
Iteration 2:   log likelihood = -13884.683  

Refining starting values:

Grid node 0:   log likelihood = -13579.247

Fitting full model:

Iteration 0:   log likelihood = -13579.247  (not concave)
Iteration 1:   log likelihood = -13573.662  (not concave)
Iteration 2:   log likelihood = -13571.294  (not concave)
Iteration 3:   log likelihood = -13569.258  (not concave)
Iteration 4:   log likelihood = -13567.897  
Iteration 5:   log likelihood = -13554.129  
Iteration 6:   log likelihood = -13554.124  
Iteration 7:   log likelihood = -13554.124  

Random-effects ordered probit regression             Number of obs    = 10,407
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    415
                                                                  avg =  743.4
                                                                  max =    973

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     =  80.42
Log likelihood = -13554.124                          Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
             legit_judi | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0165775   .0080625     2.06   0.040     .0007753    .0323798
                educlvl |   .0727078   .0149882     4.85   0.000     .0433315    .1020842
                        |
c.lnfdi_capex#c.educlvl |  -.0087733   .0023216    -3.78   0.000    -.0133235   -.0042231
                        |
                    age |   .0027179   .0009429     2.88   0.004     .0008698    .0045661
                 gender |   .0107584   .0238387     0.45   0.652    -.0359646    .0574815
                  unemp |   -.021465   .0254013    -0.85   0.398    -.0712506    .0283207
                  urban |    .106173   .0259294     4.09   0.000     .0553523    .1569938
                 income |   .0230773   .0064184     3.60   0.000     .0104974    .0356572
                intnews |   .0010368   .0222483     0.05   0.963     -.042569    .0446427
------------------------+----------------------------------------------------------------
                  /cut1 |   .0136763   .1083354                     -.1986572    .2260097
                  /cut2 |   .6589319   .1084771                      .4463208     .871543
                  /cut3 |   1.391185   .1088669                       1.17781     1.60456
------------------------+----------------------------------------------------------------
              /sigma2_u |   .1005786   .0388425                      .0471823    .2144035
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 661.12       Prob >= chibar2 = 0.0000

. margins, post predict(outcome(3)) ///
>         at(educlvl=1.40 lnfdi_capex=2.08) ///
>         at(educlvl=1.40 lnfdi_capex=8.42) ///
>         at(educlvl=4.50 lnfdi_capex=2.08) ///
>         at(educlvl=4.50 lnfdi_capex=8.42) atmeans

Adjusted predictions                                    Number of obs = 10,407
Model VCE: OIM

Expression: Pr(3.legit_judi), predict(outcome(3))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .1431042   .0195821     7.31   0.000      .104724    .1814844
          2  |   .1490491   .0201798     7.39   0.000     .1094975    .1886007
          3  |   .1825868   .0228881     7.98   0.000     .1377268    .2274467
          4  |   .1482526   .0199967     7.41   0.000     .1090598    .1874454
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                    Number of obs = 10,407
Model VCE: OIM

Expression: Pr(3.legit_judi), predict(outcome(3))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         =  35.2482 (mean)
       gender      = .4791006 (mean)
       unemp       = .4032862 (mean)
       urban       = .5821082 (mean)
       income      = .3028731 (mean)
       intnews     = .5230134 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .1431042  _b[1bn._at]
          2  |   .1490491  _b[2._at]
          3  |   .1825868  _b[3._at]
          4  |   .1482526  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0059449   .0078208     0.76   0.447    -.0093836    .0212734
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =    0.58
         Prob > chi2 =    0.4472

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |  -.0343342   .0091007    -3.77   0.000    -.0521712   -.0164972
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =   14.23
         Prob > chi2 =    0.0002

. * Legitimacy Belief: Military Control
. xtoprobit legit_mili c.lnfdi_capex##c.educlvl $C1 $C2, i(ccode)

Fitting comparison model:

Iteration 0:   log likelihood =  -10999.83  
Iteration 1:   log likelihood = -10862.409  
Iteration 2:   log likelihood = -10862.227  
Iteration 3:   log likelihood = -10862.227  

Refining starting values:

Grid node 0:   log likelihood = -10446.699

Fitting full model:

Iteration 0:   log likelihood = -10446.699  (not concave)
Iteration 1:   log likelihood = -10442.918  (not concave)
Iteration 2:   log likelihood =  -10439.25  
Iteration 3:   log likelihood = -10437.327  (backed up)
Iteration 4:   log likelihood =  -10414.07  
Iteration 5:   log likelihood =  -10414.04  
Iteration 6:   log likelihood =  -10414.04  

Random-effects ordered probit regression             Number of obs    =  9,869
Group variable: ccode                                Number of groups =     14

Random effects u_i ~ Gaussian                        Obs per group:
                                                                  min =    307
                                                                  avg =  704.9
                                                                  max =    909

Integration method: mvaghermite                      Integration pts. =     12

                                                     Wald chi2(9)     = 122.35
Log likelihood = -10414.04                           Prob > chi2      = 0.0000

-----------------------------------------------------------------------------------------
             legit_mili | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
------------------------+----------------------------------------------------------------
            lnfdi_capex |   .0652478   .0093858     6.95   0.000     .0468519    .0836436
                educlvl |   .0646933   .0174157     3.71   0.000     .0305592    .0988275
                        |
c.lnfdi_capex#c.educlvl |  -.0099267   .0026571    -3.74   0.000    -.0151346   -.0047189
                        |
                    age |  -.0031564    .001053    -3.00   0.003    -.0052202   -.0010925
                 gender |  -.0055448   .0261464    -0.21   0.832    -.0567908    .0457011
                  unemp |   .0030803   .0279602     0.11   0.912    -.0517206    .0578812
                  urban |   .1089327   .0285347     3.82   0.000     .0530057    .1648597
                 income |   .0076651   .0070853     1.08   0.279    -.0062219    .0215521
                intnews |  -.0676308   .0244826    -2.76   0.006    -.1156159   -.0196457
------------------------+----------------------------------------------------------------
                  /cut1 |   .5284298   .1326227                      .2684941    .7883654
                  /cut2 |   1.270035   .1329637                      1.009431     1.53064
                  /cut3 |   1.888316   .1335663                      1.626531    2.150101
------------------------+----------------------------------------------------------------
              /sigma2_u |   .1622754   .0625476                      .0762368    .3454146
-----------------------------------------------------------------------------------------
LR test vs. oprobit model: chibar2(01) = 896.37       Prob >= chibar2 = 0.0000

. margins, post predict(outcome(3)) ///
>         at(educlvl=1.40 lnfdi_capex=2.08) ///
>         at(educlvl=1.40 lnfdi_capex=8.42) ///
>         at(educlvl=4.50 lnfdi_capex=2.08) ///
>         at(educlvl=4.50 lnfdi_capex=8.42) atmeans

Adjusted predictions                                     Number of obs = 9,869
Model VCE: OIM

Expression: Pr(3.legit_mili), predict(outcome(3))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .0500662   .0116811     4.29   0.000     .0271716    .0729608
          2  |   .0897603   .0178354     5.03   0.000     .0548035    .1247171
          3  |   .0645626   .0141154     4.57   0.000      .036897    .0922282
          4  |   .0812745   .0165393     4.91   0.000     .0488582    .1136909
------------------------------------------------------------------------------

. margins, coeflegend

Adjusted predictions                                     Number of obs = 9,869
Model VCE: OIM

Expression: Pr(3.legit_mili), predict(outcome(3))
1._at: lnfdi_capex =     2.08
       educlvl     =      1.4
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)
2._at: lnfdi_capex =     8.42
       educlvl     =      1.4
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)
3._at: lnfdi_capex =     2.08
       educlvl     =      4.5
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)
4._at: lnfdi_capex =     8.42
       educlvl     =      4.5
       age         = 35.24683 (mean)
       gender      = .4711724 (mean)
       unemp       = .4005472 (mean)
       urban       =   .59084 (mean)
       income      = .3209038 (mean)
       intnews     = .5302462 (mean)

------------------------------------------------------------------------------
             |     Margin   Legend
-------------+----------------------------------------------------------------
         _at |
          1  |   .0500662  _b[1bn._at]
          2  |   .0897603  _b[2._at]
          3  |   .0645626  _b[3._at]
          4  |   .0812745  _b[4._at]
------------------------------------------------------------------------------

. lincom 2._at - 1._at

 ( 1)  - 1bn._at + 2._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |   .0396941   .0079751     4.98   0.000     .0240632    .0553251
------------------------------------------------------------------------------

. test 2._at = 1._at

 ( 1)  - 1bn._at + 2._at = 0

           chi2(  1) =   24.77
         Prob > chi2 =    0.0000

. lincom 4._at - 3._at

 ( 1)  - 3._at + 4._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         (1) |    .016712   .0057204     2.92   0.003     .0055001    .0279238
------------------------------------------------------------------------------

. test 4._at = 3._at

 ( 1)  - 3._at + 4._at = 0

           chi2(  1) =    8.53
         Prob > chi2 =    0.0035

. 
. 
. 
. *------------------------------------------------------------------------------*
. 
. 
. capture log close
